The national unemployment rate is 4.3%. Recent graduates are now doing worse in the job market than the average American worker, which has not been the normal pattern for people with four-year degrees. The underemployment rate for new graduates reached 42.5%, the highest since 2020.

Listen on Apple Podcasts

Graduates expect to earn $80,004 in their first year out of college. Actual starting salaries average $56,153. Applications per job posting jumped 26% while early-career postings fell 16%.

The Class of 2026 is walking into a labor market that is hiring less, paying less, and competing more than the one their older siblings entered.

Remote Work Created the Problem. AI Is About to Make It Worse.

The Federal Reserve Bank of New York published a study on June 1 attributing 64% of the rise in young graduate unemployment to remote work rather than AI. Companies are reluctant to hire inexperienced workers into roles where mentoring and on-the-job training are difficult to deliver remotely.

Unemployment among graduates under 29 rose 20% after the pandemic. Unemployment among graduates over 29 in the same fields fell slightly over the same period.

The Fed’s claim is that new workers need proximity to learn, and remote work removed that proximity. The New York Fed researchers noted their own finding may be temporary.

Stanford researchers found that early-career workers aged 22 to 25 in the most AI-exposed jobs saw employment fall 16% since late 2022 even after stripping out remote-friendly roles.

Nevertheless, the aphorism that arose during the peak-Covid remote work experience applies here to new job entrants: If you can do you can do your work remotely, so can a machine.

Jobs Are Not Disappearing. They Are Shrinking.

A May 2026 Gartner report covered by The Register describes AI’s effect on employment as hollowing out rather than elimination. AI absorbs discrete tasks within existing roles, narrows the scope of work, and compresses wages without necessarily removing the job title from the org chart.

Workers whose roles depend on judgment, context, and accountability may find AI a useful tool. Workers whose roles depend on processing, routing, summarizing, and formatting are finding themselves doing less of that work, earning wages that reflect the reduced scope, and uncertain about how the change happened.

Imperial College London and Microsoft jointly found that AI adoption paradoxically increases workplace burdens for workers who remain, as their jobs shift toward reviewing, verifying, and correcting the output of multiple AI agents. Gartner forecasts that new job categories will emerge from this transition by 2028 or 2029. That is a gap of two to three years during which the hollowing out continues without the offset.

When Every Company Automates at Once, Everyone Loses

A paper published June 3 by University of Pennsylvania researchers took at at AI FOMO (Fear of Missing Out) by herds of corporate execs as potentially harmful to their own companies interests.

Each individual company rationally reduces labor costs through AI adoption without waiting to see if the AI will actually do the jobs. When all companies do it simultaneously, consumer spending contracts, revenue bases erode, and the productivity gains from automation are collectively undermined.

Each company individually believes it is making the rational balance-sheet decision to lay off staff.

Bureau of Economic Analysis data cited in the paper shows business investment overtook consumer spending as the primary driver of U.S. GDP growth in Q1 2026. The personal savings rate fell to 3.6% in March, its lowest since late 2022. Workers are spending down savings while companies invest upward. The productivity gains are flowing into capital formation. They are not reaching household income.

The Agents Companies Are Replacing Workers With Cannot Do the Work

A CBC News investigation published June 3 pulled together the controlled performance evidence on AI agents. Carnegie Mellon’s TheAgentCompany benchmark tested 10 leading AI agent models across 175 real-world office tasks. The best-performing model completed 30.3% of tasks successfully. Gartner found AI agents produce incorrect results on office tasks roughly 70% of the time.

A major AI infrastructure company stated its agents fail to produce professionally acceptable work more than 19 times out of 20. Companies are replacing workers with these tools right now. Block’s stock jumped 20% when it announced AI-attributed layoffs.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

A University of Toronto researcher told CBC that the explanation is straightforward: “Investment in AI as an ideology has more value than keeping your workforce. That’s how capitalism works.”

Workers who remain after AI-related cuts find their jobs have changed. Their new role is reviewing, correcting, and monitoring the agents that replaced their former colleagues.

BUY NOW! Get the NEW Book that exposes the Narratives Tech uses to build its AI Empire. $4.95 flat fee for Kindle, Nook, Tablets, and Mobile. No subscription required.3.5-hr reading time.

Nine in ten graduates say they are worried AI will replace entry-level roles. Only one in three say college is preparing them to use AI in the workplace. That gap between anxiety and preparation is the condition graduates are walking into this month, in cities and towns across the country.

Also on Apple Podcasts

The mechanism producing that gap has a name: Economists call it the low-hire, low-fire labor market. A Truflation analysis published today names it more precisely: a frozen labor market. The Bureau of Labor Statistics reports 6.9 million job openings, layoffs unchanged at 1.9 million, and quits stable at 3.2 million. The headline unemployment rate sits at 4.3%. Those numbers look healthy. They are misleading.

AI does not need to fire workers to damage the labor market. It suppresses hiring. Companies retain existing staff while letting natural attrition shrink headcount. They do not replace the roles that open up.

Entry-level postings fall because AI absorbs the codifiable tasks those roles were built around. Quit rates fall because workers fear what is outside. The result is a labor market that produces stable unemployment statistics and simultaneously closes the door on anyone trying to enter it or move within it.

The frozen labor market is the statistical equivalent of the Ghost GDP pattern The AI Report documented in May: growth and stability at the aggregate level, real damage accumulating in the populations the aggregate cannot see.

BUY NOW! Get the NEW Book that exposes the Narratives Tech uses to build its AI Empire. $4.95 flat fee for Kindle, Nook, Tablets, and Mobile. No subscription required.3.5-hr reading time.

The April 2026 Challenger, Gray & Christmas report adds a specific data point to that picture. Job cuts rose 38% from March to April. AI was cited as a factor in a growing share of those announcements. Year-to-date cuts remain down 50% from the same period last year. That combination tells the same story as the unemployment statistics: fewer dramatic layoff events, but a labor market that is quietly contracting at the entry level and the middle while producing no alarm signal in the headline numbers.

The most advanced version of that dynamic is already visible in the United Kingdom. Morgan Stanley research shared with Bloomberg finds that UK companies which have used AI for at least a year reported net job losses of around 8% over the past 12 months. That is the steepest decline among comparable economies including the United States, Germany, and Japan. British firms report AI productivity gains estimated above 11%. Those gains have not been matched by job creation. Companies are banking the efficiency as margin.

BUY NOW! Get the book that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for Kindle, Nook, Tablets, and Mobile. No subscription required. 2-hr reading time.

The UK data is the leading indicator for where the United States is heading. British companies are further along the AI adoption curve. They have restructured around the technology.

The result is net job loss, not neutrality, and productivity gains that flow to shareholders rather than to the workforce that generated them.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The Class of 2026 is not graduating into a labor market that is collapsing. The unemployment statistics will confirm that next month. They are graduating into a labor market that has stopped expanding at the level where they were supposed to enter it, for reasons that will not show up clearly in any single data release.

The frozen labor market does not fire people. It simply freezes new ones out.

The company cut 700 employees, roughly 14% of its workforce, and eliminated all “pure managers” in favor of what Armstrong calls “player-coaches” who manage 15 or more direct reports while doing individual work alongside their teams.

The average American manager now oversees 12.1 direct reports, up from 10.9 in 2024. Meta’s new applied engineering team already runs a 50-to-1 employee-to-manager ratio.

Also on Apple Podcasts

Coinbase is experimenting with one-person teams where a single AI-fluent employee handles what previously required an engineer, a designer, and a product manager. Armstrong wrote that Coinbase is “rebuilding as an intelligence, with humans around the edge aligning it.”

So, the Coinbase business design goal is: AI at the center, humans at the edge, and no organizational layer in between that AI can handle itself.

The workers most immediately at risk in that design are middle managers and coordination roles. They are also the workers who have no clean path to the human-edges Armstrong describes.

A CBS News report published this week catalogues AI-attributed cuts at Pinterest, Dow Chemical, Chegg, Indeed, and Glassdoor alongside the larger announcements at Amazon, Block, and Oracle.

BUY NOW! Get the NEW Book that exposes the Narratives Tech uses to build its AI Empire. $4.95 flat fee for Kindle, Nook, Tablets, and Mobile. No subscription required.3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

Oxford Economics director Ben May told CBS that some firms are using AI as “a pretext for job cuts,” describing it as a way to “dress up layoffs as a good news story.” The AI-jobs attribution is running across an oil company, a brewer, a chemical giant, a legal education platform, and a crypto exchange simultaneously.

The pattern is not a technology sector story. It is an economy-wide story about what companies say when they restructure.

The workers absorbing these cuts are not passive. Forrester Research’s 2026 workforce forecast tracks a segment it calls “coasters,” defined as employees who do not believe their employer deserves their full effort.

That group stood at 27% of workers in 2024 and is projected to reach 28% in 2026. The mechanism is documented: workers watch colleagues cut for AI capabilities that have not materialized, see entry-level positions eliminated, and observe offshore arbitrage described as innovation.

Forrester finds that only 16% of workers had high AI readiness in 2025. Organizations are cutting staff before training the people who remain. The result is a workforce that is simultaneously smaller, less experienced, and less engaged. No AI deployment compensates for that combination at scale.

BUY NOW! Get the book that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for Kindle, Nook, Tablets, and Mobile. No subscription required. 2-hr reading time.

The sharpest contrast to all of this comes from East Asia. Governments are working from a fundamentally different premise.

A Carnegie Endowment for International Peace analysis published last week documents that South Korea, Japan, China, Taiwan, and Singapore all treat AI primarily as a solution to labor scarcity rather than a source of displacement.

Japan anticipates a 3.39 million worker shortfall in AI and robotics by 2040. China’s Ministry of Human Resources announced an employment stabilization strategy built around AI upskilling.

Korea’s 2026 National AI Action Plan funds regional competency hubs and an Inclusive Labor Transition National Strategy that includes compensation plans for AI-driven job loss.

That policy orientation produced a development reported today by the Korea Times. South Korean labor unions, employers, and government researchers are in active negotiations over what experts are calling a new “survival pact” for the AI era.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The talks center on mandatory retraining rights, stronger safety nets for displaced workers, and ethical guardrails on AI use in hiring and performance evaluation.

South Korea has the highest robot density in the world at 1,012 robots per 10,000 workers. Its government chose negotiation over displacement as its primary policy frame.

The United States has no comparable federal transition framework.

The American equivalent of the Korean survival pact conversation is currently happening in the pages of the New York Times and in Signal chats that executives delete before journalists can read them.

Sources inside frontier AI labs told the reporter, Jasmine Sun, they expressed “more extreme concern about the labor market impacts of AI in private conversation, but suddenly became optimists once I turned on the mic.”

Also on Apple Podcasts

The piece names this the San Francisco consensus.” The Consensus is a shared private belief among engineers, venture capitalists, and founders that advanced AI will displace millions of jobs, concentrate wealth in AI companies and capital owners, and produce lasting inequality.

Anthropic CEO Dario Amodei has said publicly that AI may create an unemployed or very-low-wage underclass. Sam Altman predicted in 2021 that without aggressive policy action, most people will end up worse off than they are today. These are the optimists. They are the ones willing to speak for attribution.

The investigation also documents the institutional mechanism that keeps the public conversation sunnier than the private one. When a veteran lobbyist joined OpenAI’s leadership in April 2024, the company deprioritized research that could produce unflattering findings. OpenAI has excluded public discuss of studies on the gender gap in AI’s labor market effects and on long-run economic forecasting, among others.

The company’s messaging shifted toward GDP growth stories. The progressive proposals in OpenAI’s April white paper are real ideas. Nevertheless, the absence of any specific legislation the company has committed to speaks volumes.

Get the eBook that exposes the Narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

The people with the clearest understanding of the problem are the people whose economic position makes pessimism about it most costly. For instance, Anthropic’s annualized revenue has surged to $30 billion, driven by Claude Code, an agent that automates knowledge work. The rewards for optimism, in other words, are tremendous; transparence, not so much.

Palantir CEO Alex Karp named the underlying concern plainly to a panel audience in March. “The biggest challenge to AI in this country is political unrest. If I were sitting here in private with my peers, I’d be telling them the country could blow up politically and none of us are going to make any money when the country blows up.”

Karp frames AI impact as a risk management issue rather than worker protection problem. That’s is precisely what makes the Connecticut legislation passed last Friday so significant.

Connecticut’s Senate Bill 5, now the Connecticut Artificial Intelligence Responsibility and Transparency Act, cleared the House 131-17 and the Senate 32-4. Governor Lamont confirmed he will sign it, reversing his position from 2025.

The bill requires employers to disclose when AI tools are used in hiring, promotion, discipline, or termination. It amends anti-discrimination law to confirm that deploying an AI tool is not a legal defense against a discrimination claim.

Get the eBook that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

The provision with the broadest labor market implications takes effect October 1, 2026. Any employer filing a federal WARN notice in Connecticut must disclose whether the layoffs are related to AI or technological change. The Worker Adjustment and Retraining Notification Act of 1988 is a U.S. labor law that protects employees, their families, and communities by requiring most employers with 100 or more employees to provide notification 60 calendar days in advance of planned closings and mass layoffs of employees.

That single requirement could begin to close the measurement gap the AI Labor Report has documented throughout 2026. Connecticut passed its bill by a margin of 131-17.

Meanwhile, a federal executive order directs the attorney general to challenge state AI legislation as an obstacle to national policy.

The contrast between the State and Federal approaches is sharpest when set against what Italy has already put in place.

Italy’s Law No. 132/2025 became fully operational in October 2025. The law bans fully automated employment decisions. Every employment-related decision made with AI input requires meaningful human oversight.

Before deploying any AI system in the workplace, employers must consult with trade unions. Workers receive advance disclosure when AI tools are used in hiring and evaluation and retain the right to challenge AI-driven decisions and request human review.

Italy is the first EU member state with comprehensive national AI legislation. Its union consultation requirement is the strongest worker protection provision currently in force anywhere in the developed world.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The gap between the Italian framework and the American federal posture is a gap in whose interests the regulatory architecture is designed to serve. Italy requires employers to consult trade unions before deploying AI that affects workers. The United States federal government is working to prevent states from requiring employers to disclose whether AI caused a layoff.

The San Francisco consensus says the disruption is coming. The policy question is whether the institutions making decisions about it are building the regulatory infrastructure that protects the people most exposed to the disruption — or the systems that protect the people who are accountable for the AI rollout.

Connecticut and Italy chose one answer; The federal government is choosing another.

Glassdoor chief economist Daniel Zhao identified a dynamic that goes beyond the headline number. Fewer workers are quitting, because they fear what is outside the door more than what is inside it. That reluctance to leave creates a paradox that works against the worker.

Apple Podcasts

When natural attrition slows, companies become more aggressive about pushing people out through performance management and structured layoff rounds. Workers are staying put out of fear, and their employers are treating that immobility as an opportunity to cut on the company’s terms rather than the worker’s.

The speed of the underlying change helps explain why the confidence collapse is so sharp. Cognizant’s “New Work, New World 2026” report, published this week, reassessed 18,000 tasks across 1,000 occupations and reached a finding that deserves wide attention. The average AI exposure score across all jobs is now 39%. That figure is 30% higher than Cognizant’s own forecast for where exposure would stand in 2032.

The technology arrived roughly six years ahead of schedule.

The percentage of tasks that AI cannot automate has dropped from 57% in 2023 to 32% today. That is a 25-point shift in approximately two years. The conventional assumption that AI displacement would be a slow-moving story is now contradicted by one of the largest occupational analyses in the field.

One specific finding in the Cognizant data should disrupt the dominant narrative about who is at risk. AI exposure in transportation jobs jumped from 6% to 25%. In construction, it rose from 4% to 12%. The assumption that AI is a white-collar problem and physical work is protected is losing its empirical foundation.

Get the eBook that exposes the Narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

The people who were told to learn a trade as a hedge against AI displacement are looking at a narrower window than they were promised.

The courts are beginning to grapple with what that acceleration means for workers who have already been displaced.

An appeals court in China ruled last week that a tech worker’s dismissal was unlawful after his company replaced his role with AI and offered him a reassignment at a 40% pay cut. The worker refused the demotion and was terminated.

He won at arbitration; lost the company’s appeal of that decision; and the appellate court finally affirmed: AI adoption does not automatically justify terminating a labor contract or significantly reducing a worker’s compensation.

The ruling came from one of China’s leading AI development centers, which gives it weight beyond a single employment dispute. The court’s reasoning is straightforward.

A company’s decision to adopt AI is a business strategy. That strategy does not transfer its costs to individual workers through unilateral pay cuts or forced terminations.

No court in the United States has yet reached a comparable holding. The Chinese ruling sets a precedent that American labor advocates and policymakers will watch closely.

Get the eBook that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

The human stakes of that legal question are most visible at scale in India. A CNBC analysis published this week, drawing on Bernstein research and industry data, documents that India’s top IT firms are downisizing. They hired roughly 170,000 workers in the fiscal year ending March 2026, down from a five-year average of 230,000. Tata Consultancy Services plans to hire only 25,000 fresh graduates this year, against a three-year average of 40,000.

Bernstein sent an open letter to India’s Prime Minister Modi warning of a deepening employment crisis. The letter proposed that India’s IT and BPO sector is not just an industry.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

It is the economic foundation of the country’s aspirational middle class, anchoring consumption in real estate, education, and services across dozens of cities. When IT hiring contracts, the effects ripple through the broader economy in ways that employment statistics in the sector alone do not capture.

The Glassdoor confidence collapse, the Cognizant acceleration data, the Hangzhou ruling, and the India hiring contraction are four different readings of the same underlying shift. The AI transition is arriving faster than projected. Global institutional responses — in courts, in corporate HR policy, and in national economic strategy — are running behind it.

Most Popular AI Labor Report Podcasts

Financial analysts and software engineers have seen the sharpest declines in their share of new-grad hiring. The pattern is a labor market splitting along a clear line.

AI-native technical roles and jobs requiring physical presence are growing. Generalist white-collar roles are contracting. The reasons for that shift are arriving in waves.

Also listen on Apple Podcasts

Meta confirmed this week that its first round of layoffs begins May 20, cutting approximately 8,000 employees across every major division of the company, including Facebook social, Reality Labs, recruiting, sales, and global operations.

The company is also canceling 6,000 open positions it had planned to fill.

The effective reduction is 14,000 positions. A second round of cuts is planned for the second half of the year.The cuts are structural, not performance-based.

Meta is reorganizing its teams into AI-focused units. New internal role categories have been created: “AI builder,” “AI pod lead,” and “AI org lead.” Engineers from across the company are being moved into the Applied AI organization. The workers being cut are not underperformers. They are people whose roles no longer fit the organizational design Meta is building around AI.

Meta generated $201 billion in revenue in 2025, up 22% year over year. Its 2026 capital expenditure guidance runs as high as $135 billion. The company is not cutting from financial weakness.

Instead, it is converting payroll budgets into AI infrastructure budgets, and it is doing so while its stock trades near record levels. For workers, the financial health of their employer provides no protection against structural reorganization.Meta is one data point in a larger pattern, though.

Updated tracking from Layoffs.fyi and Nikkei Asia puts the 2026 tech sector total above 95,000 job cuts across 247 layoff events, averaging 882 workers per day. Nearly half of those cuts are attributed to AI or workflow automation. The cumulative total since 2020 is approaching 900,000. The cuts are arriving ahead of the operational justification for them.

Get the eBook that exposes the Narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

Cognizant’s chief AI officer told Nikkei Asia this week that real productivity gains from AI are still six to twelve months away for most companies. That gap between the cuts and the demonstrated productivity gains is where the most serious long-term damage is accumulating.

A Fortune piece this week, drawing on MIT research, warns that companies eliminating junior roles to extract short-term AI efficiency gains may be destroying their own future leadership pipelines. The mechanism is straightforward.

Entry-level positions are where workers develop the tacit knowledge that eventually makes them valuable at senior levels. AI can handle the codifiable tasks those positions require.

It cannot transfer the judgment and pattern recognition that come from years of doing the work.Companies that stop hiring junior workers today will face a leadership gap in five years that they cannot quickly or cheaply reverse.

Get the eBook that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

IBM reached that conclusion and responded by tripling its entry-level hiring in 2026 against the prevailing industry trend. The question is whether the companies eliminating entry-level positions now are making a short-term efficiency calculation that will cost them more than it saves.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The Gusto hiring data and the MIT warning are two sides of the same problem. The labor market is reorganizing around a rationale that rewards physical presence, AI fluency, and deep domain expertise while compressing the middle.

Workers building careers in that middle need to understand the reorganization clearly. The companies making the cuts are counting on the fact that most of them do not.

The entire justification workers have been given for their displacement is that AI is cheaper and more efficient than people.The workers who lost jobs to that cost assumption have no mechanism for recourse while the economics get worked out.

Amazon made the most consequential single announcement of the week for middle-American workers. The company launched Connect Talent, an agentic AI system that conducts voice job interviews around the clock, scores candidates on competency, and prepares recruiter notes entirely without human intervention. Amazon built the system to handle 250,000 seasonal hires.

It is now selling that system to retailers, manufacturers, logistics operators, and hospitality companies. The human job interview is now an optional feature. The human recruiter is the first professional role Amazon has automated and packaged for sale at scale to the industries where most middle-American service workers are employed.

Also listen on Apple Podcasts

The macro picture around those two stories is built on a set of numbers that do not fit together easily. The BEA reported Q1 2026 GDP growth of 2.0%. Jobless claims dropped to 189,000, the lowest level since 1969. Corporate earnings across the Magnificent Seven were strong.

Also, the same quarter produced the largest single-week cluster of AI-attributed layoff announcements in corporate history, with 92,000 tech workers cut so far in 2026 and nearly half of those cuts attributed to AI.

The economy is growing. Workers are being cut. The headline statistics look healthy because the workers absorbing the displacement are contractors, gig workers, and early-exit buyout takers — populations the unemployment system was never designed to count.

Get the eBook that exposes the Narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

Available on Gumroad

The people building artificial intelligence did not invent their ideas. They inherited them.

The Citrini Research bit of speculative fiction that moved markets in February called this “Ghost GDP”: growth that shows up in national accounts but never circulates through the people who used to earn it. This week’s data made that speculative scenario visible in actual earnings filings and government statistics, simultaneously.

Meanwhile, Anthropic’s March 2026 labor market research measured actual AI task coverage in real workplace deployments. Computer programmers show 74.5% observed task coverage. Customer service representatives are at 70.1%. The most exposed workers are more likely to be female, over 40, more educated, and better compensated.

The entire “get a degree and move into knowledge work” strategy a generation of middle-American workers followed is contradicted by the actual usage data from the largest AI deployment in the world.

Regardless, workers using AI daily are producing more output and losing underlying competency. A January 2025 controlled study found that participants who received AI assistance during training performed substantially worse on subsequent independent assessments.

Senior workers who built expertise before AI became ubiquitous can use AI as an augmentation layer because they understand their domain well enough to evaluate its output.

Get the eBook that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

Junior workers building careers on top of AI from day one are building on a foundation they have never been required to develop independently.

What the week’s stories mean together

Three trends are now running simultaneously and reinforcing each other: The first is that AI costs more than projected, organizational cuts are made in anticipation of savings that have not yet arrived, and workers bear the cost of that gap.

Further, the measurement architecture — unemployment claims, GDP, monthly jobs reports — was built for a workforce that no longer describes the majority of American workers. The contractor economy absorbs displacement in silence.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

And then, the workers most at risk are precisely the workers who followed the advice they were given: get educated, move into knowledge work, build credentials.

More practical advise would be: do what AI cannot do for at least a generation into the future.

The essay was speculative fiction. It was also the most widely read piece of financial writing this year, and the reason it resonated is visible in this week’s data.

The companies that reported earnings this week posted strong results. Meta raised its full-year AI capital expenditure guidance to between $125 and $145 billion while confirming 8,000 layoffs beginning May 20. The Bureau of Economic Analysis (BEA) advance estimate released Thursday shows the economy grew at an annualized rate of 2.0% in Q1 2026, a meaningful rebound from Q4 2025’s weak 0.5% expansion.

Also listen on Apple Podcasts

A significant driver of the growth was from construction of data warehouses.

The economy is growing. Corporate profits are strong. And the same quarter produced the largest single-week cluster of major AI-attributed layoff announcements in corporate history.

The Ghost GDP speculation describes something real: the gap between aggregate economic growth and the experience of workers inside that growth. The GDP number does not contradict the job displacement story. It sharpens it.

Growth driven by AI infrastructure investment and productivity gains flowing to shareholders does not automatically reach the worker whose coordination role just became unnecessary.

Get the eBook that exposes the Narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

That gap is precisely what economist Claudia Sahm has been warning about for months. Sahm, the inventor of the Sahm Rule recession indicator, describes three forces converging on the same population in the same quarter: AI displacement of white-collar roles; federal workforce reductions under DOGE; and broader economic uncertainty suppressing private-sector hiring.

Federal workers pushed into early retirement or layoffs are now competing for private-sector jobs in a market where white-collar hiring has been contracting on net for three consecutive years.

Both forces are hitting educated professionals with credentials who assumed knowledge work was structurally protected.

Sahm does not describe this as a crisis that has fully arrived. She describes it as a convergence that 2026 could make visible in the labor data for the first time. The GDP growth number makes that convergence harder to see, not easier. A 2.0% growth rate absorbs a great deal of localized professional pain without registering it.

The jobless claims number released Thursday morning makes the same point from a different direction. Initial claims dropped to 189,000, the lowest level since 1969. Economists had forecast 212,000.

Claims, though, count W2 employees who lose jobs and file for benefits. The job displacement wave documented this week in The AI Labor Report runs significantly through channels that claims data cannot reach: contractors whose engagements end quietly, entry-level positions that go unfilled rather than eliminated, freelancers whose work volume compresses without a termination, and senior workers who accept a voluntary buyout and exit on their own terms.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Meta’s 8,000 layoffs begin May 20. Those workers have not filed claims yet.

GDP growth, jobless claims, and Big Tech earnings all look reassuring in isolation. Together, they describe an economy that is producing growth and cutting workers at the same time, in a way that the standard measurement architecture was not designed to capture.

That is not a contradiction.

The Ghost GDP essay was fiction. The economic dynamic it described is documented in this week’s data releases, across four separate government and corporate sources, on the same Thursday morning.

Audio Production: Eleven Labs

Previous studies estimated AI’s labor impact by asking what AI could theoretically do. Anthropic’s researchers asked a different question: what is AI actually doing in professional settings right now?

Also listen on Apple Podcasts

They combined occupational task data, capability estimates from prior research, and real usage data from Claude deployments across enterprise environments. The result is a metric they call “observed exposure” — a measure of how much of a job’s actual task load AI is already handling today.

Computer programmers show 74.5% observed task coverage. Customer service representatives follow at 70.1%. Data entry operators are at 67.1%. The data are measurements of what AI is doing right now in real workplaces, on real Tuesday afternoons, for real employers.

The report documents an early labor market signal that has not yet appeared in unemployment statistics. The unemployment rate for workers in highly exposed occupations shows no spike since ChatGPT’s launch in late 2022.

The hiring rate for workers aged 22 to 25 entering those same occupations has slowed by approximately 14%. The signal is appearing in hiring before it appears in unemployment. Workers are not being fired at unusual rates.

Get the eBook that exposes the Narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

They are simply not being replaced when they leave, and the entry-level pipeline is quietly closing.

The Yale Budget Lab’s April 2026 update, which incorporates Anthropic’s February usage data, reaches a consistent conclusion.

Occupational patterns remain within historical ranges.

Unemployed workers are in occupations with roughly 25 to 35 percent AI task exposure; the same share as employed workers. The aggregate unemployment statistics show no AI displacement signal yet.

In other words,the labor market headline numbers are stable.

Both observations are true simultaneously: The aggregate statistics show stability; the corporate earnings calls last night offered something else entirely.

Meta raised its full-year AI capital expenditure guidance to $125 to $145 billion on the same call where it confirmed 8,000 layoffs beginning May 20. The company’s earnings announcement cited “higher component pricing” as a primary driver of that upward revision — a direct reference to tariff-driven hardware cost increases.

Alphabet, Microsoft, Meta, and Amazon are collectively spending close to $700 billion on AI infrastructure this year. Each of those companies confirmed last night, on earnings calls, on the record, to Wall Street analysts, that the spending and the cutting are two sides of the same strategic decision.

Workers are being cut to fund infrastructure that is itself getting more expensive because of trade policy.

Get the eBook that examines how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

The Anthropic report’s most counterintuitive finding ties all of this together. The workers with zero observed AI exposure in the data include cooks, motorcycle mechanics, lifeguards, bartenders, and dishwashers.

The workers with the highest exposure are educated, experienced, and well-compensated knowledge workers. The most exposed group is more likely to be female, over 40, more educated, and better paid than the zero-exposure group.

The entire “AI will take blue-collar jobs first” manufacturing narrative is contradicted by the actual usage data from the largest AI deployment in the world.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The practical implication for middle-American service workers is that the people ahead of them in the layoff queue are the white-collar professionals they were told to aspire to become.

The aggregate unemployment statistics are stable. The hiring pipeline for young workers is narrowing. The companies doing the cutting are raising their AI spending guidance on the same earnings calls. And the workers most exposed are the ones nobody expected.

It’s all part of the same story.

The company launched Connect Talent, an agentic AI system that conducts voice job interviews around the clock, scores candidates on skills, and prepares recruiter notes.

All without without human intervention.

Names and resumes are stripped from the process. Recruiters receive anonymized competency scores and transcripts. The face-to-face job interview is now an optional feature.

Also listen on Apple Podcasts

The company is now packaging the system to sell to retailers, manufacturers, logistics operators, and hospitality companies — every industry that hires large numbers of workers at once.

The human recruiter is the first professional role Amazon has fully automated and then offered for sale at scale to other industries. The product is live today.

The announcement arrives on the same morning that a more uncomfortable AI economics story is breaking across the business press.

Axios reported that Nvidia’s VP of Applied Deep Learning, Bryan Catanzaro, told reporters that the cost of compute for his team now exceeds the cost of employees. The company whose chips power most of the AI economy.

Uber’s CTO told The Information that the company burned through its entire 2026 AI budget on token costs before the year reached its midpoint. The culprit was coding tools.

Get the eBook of the Substack series that dissects the narratives Tech uses to build its AI Empire. $4.95 SPECIAL SALE PRICE for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

The assumption behind most AI-driven workforce reduction has been that software is cheaper than people. That assumption is now under direct pressure from the people inside the industry who know the costs best.

Human payroll is a fixed and predictable expense. AI costs, though, scale with usage. Every query, every agent action, and every model call generates an invoice. Companies that made permanent workforce decisions based on projected AI cost savings are discovering that the operational math is more complicated than the original pitch.

An AI and finance professor at the Swiss Institute of Artificial Intelligence told Fortune the situation reflects a “short-term mismatch” between current compute costs and the productivity gains companies projected when they made those cuts. The workers who lost their jobs to that mismatch have no mechanism for recourse while the labor market responds to economic realities.

The economic context around all of this shifted this morning. The Bureau of Economic Analysis released its advance estimate for Q1 2026: U.S. GDP contracted 0.3%. *

Into that landscape steps a story that has received less attention than it deserves. Fortune today profiles Clara Shih, the former head of AI at both Meta and Salesforce.

She describes watching AI agents outperform her top human workers on measurable tasks in real enterprise settings. She describes the experience as having “radicalized” her.

Get the eBook compilation of Substack articles examining how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

She has since left corporate AI work and built a nonprofit aimed at helping Gen Z workers find jobs before those jobs disappear.

Shih’s account matters because was a senior executive describing what she saw inside two of the largest technology companies in the world: AI agents beating experienced human workers on real work products in direct comparisons.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The pattern running through today’s news is that AI technology is advancing faster than anyone — including its builders, investors, employers, and workers — had planned for.

*CORRECTION: The BEA's advance estimate released Thursday, April 29, 2026 actually shows the economy grew at an annualized rate of 2.0% in Q1 2026, rebounding from Q4 2025's weak 0.5% expansion, due mostly to investment in data center construction.

Google engineers are no longer prompting AI to help them write code. They are directing autonomous AI agents that plan, generate, test, and iterate across entire projects.

Pichai described a complex code migration that agents and engineers completed together six times faster than the same work would have taken a year ago with engineers alone. Google has also tied AI adoption goals to engineer performance reviews. Using these tools has become a job requirement.

The direct question for every software professional reading this: if the world’s largest engineering organization no longer needs engineers to write most of its code, what does it need them for — and how many of them does it need?

Get the eBook of the Substack series that dissects the narratives Tech uses to build its AI Empire. $9.95 flat fee for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

The Google disclosure lands alongside a body of research that is now documenting a quieter and more insidious workplace transformation. Researchers and employers have given it a name: AI deskilling.

The pattern is consistent across software engineering, financial analysis, legal research, and medical diagnostics. Workers who rely on AI for core job functions produce more output in less time. Their productivity metrics look strong. Underneath those metrics, their ability to perform the same functions independently is deteriorating.

A January 2025 study found that participants who received AI help during training performed substantially worse on subsequent independent assessments than those who worked through tasks without assistance. The AI completed the task. The worker never learned how to do it.

A Microsoft Research controlled experiment found that developers using GitHub Copilot completed coding tasks 55.8% faster. A separate Microsoft and Carnegie Mellon study of 319 knowledge workers found that AI tool use reduces critical engagement and raises concerns about long-term reliance and diminished independent problem-solving.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Andrea Olson, writing in Inc. Magazine, frames the structural problem clearly: productivity dashboards stay green while the foundation erodes. The dashboards track output. They do not track whether a financial analyst still understands the assumptions behind the model AI just built for her. They do not flag that a junior lawyer can no longer construct a legal argument from scratch because he has been reviewing AI-generated briefs for two years.

Senior workers who built deep expertise before AI tools became ubiquitous retain that knowledge for now. They can use AI as a genuine augmentation layer because they understand their domain well enough to evaluate, correct, and direct AI output.

Junior workers are building their careers on top of AI from day one. Controlled studies suggest many are building on sand. When the tools change, workers without underlying domain expertise will find themselves stranded.

Also listen on Apple Podcasts

The deskilling problem connects directly to a structural shift in how AI companies are now selling their products. A 2026 industry analysis documents the emergence of what analysts are calling vertical AI: companies selling the outcome of labor rather than tools to help humans work.

Insurance, legal, logistics, and healthcare administration are the primary target sectors. The AI coding tools market alone reached $12.8 billion in 2026 revenue, more than double the $5.1 billion generated in 2024.

So, the dollar amount a company previously paid a human worker becomes the revenue target for a vertical AI startup. The labor budget becomes a software subscription.

MIT Technology Review’s roundup of the ten most consequential AI trends of 2026 identifies something the displacement data alone cannot capture: a global backlash is building.

Get the eBook compilation of Substack articles examining how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

Labor unions, artists, conservative legislators, and progressive activists are converging on AI regulation from different directions and for different reasons. Small regulatory wins are accumulating.

Connecticut passed worker-protection legislation this month. Colorado’s AI law takes effect in June. The federal government has produced a framework with no binding obligations, and states are filling the vacuum.

The backlash is the political story running underneath all the economic data. Workers are sensing displacement before the aggregate statistics confirm it.

A Mercer survey of 12,000 workers and executives globally found that concern about AI-driven job loss rose to 40% in 2026, up from 28% in 2024. Ipsos data on U.S. public opinion finds that almost three in four Americans believe the government should act to prevent AI-induced job losses.

The gap between that sentiment and the current state of federal policy is the space where the next few years of labor politics will be challenged.

The company announced last Thursday that roughly 8,750 U.S. workers — about 7% of its American workforce — are eligible for a voluntary retirement program. The offer targets employees at or below the senior director level whose combined age and years of service add up to 70 or higher. Eligible workers and their managers will receive details on May 7.

The formula has a name now: the Rule of 70. It targets mid-to-late career employees, often the people who carry the most institutional knowledge in a company. In most previous periods of technological transition, that knowledge was precisely the thing that protected those workers. In an environment where AI systems can process decades of documentation in seconds, that protection is dissolving.

Get the eBook of the Substack series that dissects the narratives Tech uses to build its AI Empire. $9.95 flat fee for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

The program is voluntary, which gives Microsoft legal and reputational cover that a direct layoff round would not. It is also, by the analysis of several HR observers, the most efficient way to reduce a high-salary workforce without age-discrimination exposure. Other large employers are almost certain to follow the same template.

The Microsoft buyout is the most visible single story of the past week. The broader numbers behind it are larger and grimmer. According to data tracked by Layoffs.fyi, over 92,000 tech workers have been laid off so far in 2026. The cumulative total since 2020 stands at nearly 900,000.

Of the 78,557 tech cuts recorded from January through early April, 47.9% have been attributed to reduced need for human workers because of AI and workflow automation, according to Nikkei Asia’s tracking.

Cognizant’s chief AI officer told Nikkei that real productivity gains from AI are still six months to a year away for most companies. The cuts are arriving ahead of the operational justification for them.

Get the eBook compilation of Substack articles examining how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

That gap between the cuts and the demonstrated productivity gains shows up most visibly at the entry level. The Stanford 2026 AI Index found that employment among software developers aged 22 to 25 has fallen nearly 20% since 2024, even as older colleagues’ headcount continues to grow. The same pattern is appearing in customer service and other AI-exposed occupations.

The mechanism for the disruption is straightforward. AI now handles the codifiable tasks that entry-level jobs were built around: drafting standard documents, running routine analysis, producing basic reports.

Employers are retaining experienced workers and letting AI absorb the on-ramp that younger workers used to climb. The headline unemployment rate holds steady. The entry-level job market is contracting.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The Stanford data adds precision to something The AI Labor Report has tracked all year: the career ladder is being pulled up from the bottom, and the official statistics are not built to show it.

Workers in companies that have already adopted AI are absorbing this reality in real time. A new Gallup survey of 23,717 U.S. employees found that 23% of workers in AI-adopting organizations say it is very or somewhat likely their job will be eliminated within the next five years due to AI or automation. Among all U.S. workers, that share is 18%.

Gallup also found that 27% of employees in AI-adopting organizations say their workplace has changed in disruptive ways over the past year, compared to 17% in organizations that have yet to adopt AI. The disruption is real and measurable.

So is the productivity improvement: 65% of workers in AI-adopting organizations say AI has improved their efficiency.

The combination of higher disruption and higher productivity in the same organizations is the central tension of the current moment. The companies that are furthest along in AI adoption are producing more output per worker and cutting workers at the same time.

The workers who remain are doing more. The workers who leave are being converted into infrastructure budgets.

The Microsoft Rule of 70, the 92,000 layoffs, the collapsing entry-level hiring pipeline, and the Gallup anxiety numbers are all expressions of the same underlying pattern: AI is raising the output ceiling while narrowing the floor for who gets to stand on it.

It runs on a list of work-related apps and websites and captures screen content for context. Meta’s stated purpose is to improve AI performance on specific tasks where models still struggle, such as navigating dropdown menus and using keyboard shortcuts.

Also listen on Apple Podcasts

Meanwhile, Meta is preparing to cut as much as 20 percent of its workforce, with the first layoffs reportedly set to begin in May. Workers are being recorded doing their jobs in order to teach systems the work those same workers currently perform.

The company has committed up to $135 billion in capital expenditure for 2026, most of it directed toward AI. The workers generating the training data are not receiving additional compensation for it. There is no reported opt-out mechanism.

Every office worker reading this should ask one question: is my employer doing this too?

The scale of what employers are planning clarifies why that question matters. A working paper published by the National Bureau of Economic Research, drawing on a survey of nearly 750 corporate executives conducted with Duke University and the Federal Reserve Banks of Atlanta and Richmond, projects AI-driven job cuts in 2026 at approximately 502,000 roles. The 2025 figure was roughly 55,000. The projected 2026 number is nine times higher.

Get the eBook of the Substack series that dissects the narratives Tech uses to build its AI Empire. $9.95 flat fee for PDF or ePUB formats; no subscription required. 3.5-hr reading time.

The people building artificial intelligence did not invent their ideas. They inherited them.

The researchers also document what they call a productivity paradox: executives perceive AI’s productivity gains as larger than the data actually shows. Companies are cutting based on expectations that have not yet materialized in their revenue figures.

That gap between expectation and reality shapes the bind workers currently face. A survey of 2,400 knowledge workers across the U.S., the U.K., and Europe — conducted by enterprise AI firm Writer and research firm Workplace Intelligence — found that 29 percent of employees admit to actively sabotaging their company’s AI rollout. Among Gen Z workers, that figure rises to 44 percent.

The most common reasons cited are fear of job loss and frustration with tools that do not work as promised. The sabotage takes forms ranging from refusing to use approved tools to entering proprietary information into public AI platforms without authorization.

However, workers who resist AI adoption face termination risk: 60 percent of executives surveyed said they are considering cutting employees who refuse to adopt AI tools.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Workers who embrace the tools and demonstrate their capabilities face a different risk — they may be accelerating their own replacement. Seventy-seven percent of executives said workers who remain AI-resistant will be excluded from promotions and leadership consideration.

The person building the technology making the most widespread use of it has his own assessment of where this leads. Anthropic CEO Dario Amodei warned at Davos in January 2026 that AI’s reach across cognitive work is categorically different from earlier automation waves.

Factory workers displaced by industrial automation could, in theory, move into service or office roles. If AI is moving into office roles simultaneously, Amodei argued, the adjacent sector that absorbed earlier displacement does not exist. “The technology is not replacing a single job,” he said, “but acting as a general labor substitute for humans.”

So, one of the people most responsible for building the technology is the one making the case that the safety valve is gone.

Get the eBook compilation of Substack articles examining how the AI invasion already happened. You just weren’t invited. $9.95 flat fee for PDF or ePUB; no subscription required. 2-hr reading time.

Comments

It prohibits employers from deploying AI tools that produce discriminatory outcomes. It funds an AI literacy program for workers and small businesses and creates a state regulatory sandbox where companies can test new AI tools under oversight.

The bill’s sponsor, Senator James Maroney, was direct about the federal vacuum it fills: “We can no longer wait for DC to act.” The bill now moves to the House before a May 6 deadline.

Also listen on Apple Podcasts

Here’s why the timing matters: The White House published a “National AI Legislative Framework” in March calling for a unified federal approach to AI regulation. The document contains no binding obligations on employers. It signals that the administration wants to preempt state laws while producing no federal replacement.

Connecticut’s Senate passed its worker protection bill two days after that framework circulated widely. The collision between a federal preemption posture and accelerating state legislation is now the central regulatory story in AI and employment.

Workers in Connecticut may know by May 6 whether their employer must tell them an algorithm screened them out. Workers in most other states have no such right.

The urgency behind Connecticut’s bill becomes clearer when set against what is happening to the managers who are supposed to be implementing AI at the team level.

Fortune reported on April 7, drawing on Bureau of Labor Statistics data, that the average American manager now oversees 12 direct reports. That figure has nearly doubled since Gallup began tracking it in 2013.

Companies have spent three years gutting middle management and leaving whoever survives with a dramatically larger portfolio of people. Gartner found that 75% of HR leaders believe managers are already overwhelmed by their expanding responsibilities.

Yesterday’s AI Labor Report established that workers are 8.7 times more likely to say AI has transformed their work when their direct manager actively supports adoption. The managers capable of driving that transformation are being eliminated or buried under workloads that make mentorship and coaching impossible.

A Harvard Business Review piece published this month, co-authored by Wharton researchers, puts a name on the resulting organizational failure. Senior executives experience AI as a strategic advantage. Middle managers confront its actual failures inside real workflows, under real constraints, without adequate support or time.

AI initiatives stall because the people charged with making them work see a fundamentally different reality than the people who approved the budget.

The Wharton authors are direct: companies are treating AI adoption as a top-down mandate and skipping the operational burden in the middle. The result is billions in investment producing the outcome Gallup The AI Labor Report discussed this week — widespread individual use, almost no organizational transformation.

Into that gap steps Jack Dorsey. In a post published April 1 with Block’s lead independent director Roelof Botha,

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Dorsey argued that AI should replace middle management entirely.

The post describes AI handling coordination, task assignment, project tracking, and real-time business insights — the core functions that management layers have historically performed.

Block cut 40% of its workforce before publishing the piece. Dorsey’s post is one of the first explicit executive statements treating middle management elimination as an organizational design goal rather than an unfortunate side effect of restructuring.

The three stories are connected.

The managers who would implement AI humanely are being removed.

The executives who remain are treating AI as a mandate from above.

The workers who bear the cost of both decisions are now waiting to see whether state legislatures will step in with the transparency and protections that the federal government has declined to provide.

So the AI deployment battleground narrows from ructions between labor-and-industry to workers-and-management.

Check out the Future Forwarded Collections of its most popular Substack series. Each collection is $9.95 flat fee, in PDF or ePUB formats. No subscription required. Each collection takes about two- to three-hours to read.

Meanwhile, only 12% of workers say AI has fundamentally transformed how their organization actually operates. Workers are using AI tools individually, for their own tasks, in their own workflows.

Their employers have not figured out how to translate that individual use into organizational change. A separate survey of nearly 6,000 executives by the National Bureau of Economic Research found that 89% report no measurable effect of AI on their company’s labor productivity over the past three years.

Gallup’s own analysis points to the reason. Employees are 8.7 times more likely to say AI has genuinely changed how their work gets done when their direct manager actively supports and champions AI adoption.

The barrier is managerial. Manager engagement in the U.S. has fallen from 31% in 2022 to 22% in 2025. The people structurally positioned to drive AI adoption at the team level are the most disengaged group in the American workforce.

The gap between adoption and transformation matters because it shapes who bears the cost of AI’s disruption and who captures its gains. The workers absorbing the most pressure right now are young workers trying to enter the labor market.

Anthropic’s own research team published findings in March showing that hiring into AI-exposed jobs fell 14% for workers aged 22 to 25 since ChatGPT’s release in late 2022. Unemployment for that group is flat.

The hiring door is simply closing. Young workers are being excluded from the front end of the career ladder at precisely the moment when AI is absorbing the entry-level tasks those jobs used to require.

The emotional response to that exclusion is showing up in data. A Gallup survey released April 9, conducted for the Walton Family Foundation, found that 31% of Gen Z workers now report outright anger toward AI, up from 22% a year ago.

Excitement fell 14 points. Hopefulness fell 9 points.

Weekly AI use among Gen Z (those born from 1997 to 2012) grew only 4 points over the past year, well below the pace of prior years. The generation widely described as AI’s natural beneficiaries and most fluent early adopters is turning against the technology. The reason is straightforward.

They are watching it close the doors they expected to walk through.

The Stanford 2026 AI Index, released April 13, documents what is opening on the other side of those closed doors.

Demand for “Agentic AI” skills in U.S. job postings jumped 280% in a single year. These are skills related to managing and directing autonomous AI systems, not simply using AI tools.

AI skills now appear in 2.5% of all U.S. job postings, up 55% from the prior year. That hiring is concentrated in California, Texas, and New York. Singapore leads globally at 4.7% of postings.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The Stanford data also captures a broader divide that runs through all of this. Among AI experts, 73% expect AI to have a positive impact on how people do their jobs. Among the general public, only 23% agree.

That 50-point gap appears consistently across the economy and healthcare as well. The U.S. recorded the lowest trust in its own government to regulate AI of any country surveyed, at 31%.

So, 73% of experts expect AI to improve how people do their jobs; 31% of the people those jobs were supposed to belong to are angry.

The gap between techie expectations and worker realities demarcates the battleground for the next great political fight between labor and industry.

A compilation of the Substack articles examining how the invasion already happened. You just weren’t invited. $9.95 flat fee for the bundle (PDF, ePUB), no subscription required. 2-hr reading time.

No call from a manager. No meeting with HR. Just a brief email and a locked screen.

The story spread across Reddit, LinkedIn, and Blind within hours, as roughly 30,000 employees globally — about 12,000 of them in India — tried to make sense of what had just happened. The specific detail that traveled furthest was the lockout itself.

Workers described finishing their morning coffee and finding themselves locked out of the tools they had used that morning.

For anyone who has ever held a job and worried about losing it, that detail feels different than reading a headcount number.

It seems Oracle borrowed $58 billion in new debt over two months to fund data center construction for AI clients, including OpenAI, Meta, and Nvidia. The layoffs are expected to free up $8 to $10 billion in cash flow, according to TD Cowen analysts.

Oracle is not cutting because AI made those workers redundant today. Oracle is cutting because it took on massive debt and needed the payroll budget to service it. The workers are not being replaced by machines. They are being converted into debt payments.

Underneath the headline numbers, a Brookings Institution study published in January draws a distinction that most AI coverage misses entirely.

An important aspect of the the study is the concept of “AI exposure:” How much of your job AI can technically perform. The study relates AI exposure to “adaptive capacity:” a person’s actual ability to find new work if they lose theirs.

Most of the 37 million highly exposed U.S. workers have reasonable adaptive capacity. They have savings, transferable skills, and labor markets dense enough to absorb a transition.

But 6.1 million workers have high exposure and low adaptive capacity simultaneously. Those workers are primarily in clerical and administrative roles. They are secretaries, HR clerks, medical assistants, and tax preparers. Roughly 86 percent of them are women.

They tend to be older, have limited savings, and live in smaller labor markets in the Western Rocky Mountain states and Midwest where retraining options are thin.

A follow-up Brookings analysis published last week sharpens the picture further. Of the workers with both high exposure and low adaptive capacity, 67 percent are: workers who built their skills through trade, military service, or on-the-job experience rather than a four-year degree. These are workers who used entry-level and mid-level positions as rungs on a ladder toward economic stability.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

AI is not just threatening their current jobs. It is eliminating the pathway itself. The routes that allowed a skilled worker without a college degree to move up through an organization are narrowing precisely as the jobs at the bottom of those routes disappear.

The population most at risk, in other words, is older women in administrative roles in smaller American cities who built their careers through experience rather than credentials.

That is a specific, recognizable group of people. They are the workers least equipped to absorb the cost and least visible in the aggregate statistics that currently show a stable labor market.

A compilation of the Substack articles examining how the invasion already happened. You just weren’t invited. $9.95 flat fee for the bundle (PDF, ePUB), no subscription required. 2-hr reading time.

The company said the savings will fund AI infrastructure investments. The news fits a market pattern: Snap’s stock rose on the announcement, which means investors are rewarding the conversion of payroll into AI spending. That financial incentive now applies to every publicly traded company watching Snap’s share price.

The broader jobs picture looks calmer on the surface. The Bureau of Labor Statistics reported that the U.S. economy added 178,000 jobs in March, and the national unemployment rate edged down to 4.3 percent. Those headline numbers are healthy by recent standards.

Underneath them, though, the Federal Reserve Bank of New York found that workers aged 22 to 27 face an unemployment rate of 5.6 percent, well above the national figure.

The reason is structural. AI now handles the codifiable tasks that entry-level jobs used to provide: drafting basic documents, producing standard reports, running routine analysis.

Employers are keeping experienced workers and letting AI absorb the on-ramp that younger workers traditionally climbed. The headline unemployment rate looks fine. The entry-level job market is a different story entirely.

IBM is reading that situation as a strategic risk. The company has tripled its entry-level hiring in 2026, bucking the prevailing industry trend.

IBM argues that eliminating junior roles produces short-term savings but destroys the pipeline that generates experienced workers and managers five years down the road.

A company that stops hiring entry-level workers today will face a serious leadership gap in the future. IBM is also arguing that AI tools still need human direction, judgment, and context.

The tripled hiring is a bet that applied AI literacy combined with human judgment produces better outcomes than AI operating without new staff.

Whether other major employers follow that reasoning, or continue optimizing for immediate cost reduction, will shape the career prospects of an entire generation.

The European Central Bank published research this week drawing on surveys of more than 5,000 euro-area firms. The finding is that companies actively investing in AI are 4 percent more likely to add headcount than their non-AI peers.

The employment benefit shows up primarily at small firms. Large firms show a neutral effect. The ECB’s explanation is that AI investment tends to accompany business growth, and growing businesses hire people.

That dynamic holds in Europe’s current environment. The important caveat is the phrase “for now.” The ECB economists are describing short-term outcomes in firms that are still early in their AI adoption. The same firms that are hiring today may face different pressures once AI capabilities mature further.

The most striking labor story of the week came out of China. A Shanghai engineer built a tool called Colleague Skill, reported today by MIT Technology Review.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The tool works by ingesting an employee’s work chat history and internal documents. It then produces a detailed profile of how that person does their job, including their reasoning patterns, communication style, and decision-making approach.

The stated purpose is to allow a company to automate that employee’s role. The tool went viral on Chinese social media, initially as dark humor. Workers joked about automating their coworkers before their coworkers automated them. The debate that followed turned serious quickly.

Legal scholars pointed out that a work chat history may belong to the employer, but the judgment, tone, and professional instincts embedded in it belong to the person. Colleague Skill forces a question that no labor law in any country has yet answered cleanly: when AI learns how you think from materials you produced at work, who owns that knowledge?

A compilation of the Substack articles examining how the invasion already happened. You just weren’t invited. $9.95 flat fee for the bundle (PDF, ePUB), no subscription required. 2-hr reading time.

It is the central fact of the labor market in 2026. On one side, Morgan Stanley published a careful analysis this week tracing five major innovation waves across 250 years.

Chief U.S. Economist Michael Gapen is explicit that those waves “are disruptive, capital-intensive and often volatile. They can displace workers, concentrate gains early and provoke political backlash. But over time, they raise productivity, restructure labor markets, expand output and, when institutions adapt, improve living standards broadly.”

The phrase “when institutions adapt” is doing enormous work in that sentence. The history bears the caveat out. Real wages in Britain barely moved during the first 80 years of industrialization, a period Acemoglu and Johnson document in Power and Progress as a sustained collapse in working conditions for the people inside the transition.

Child labor, 14-hour shifts, and unsafe factories were the daily reality, not a deviation from it. The right to organize, factory safety laws, and meaningful limits on working hours took another 50 years to win after that.

The “long run” in Morgan Stanley’s framing was, for actual workers, roughly a century and a half of paying the cost for gains that would eventually reach their great-grandchildren.

On the other side of the argument, three of the most influential labor economists alive, Daron Acemoglu, Simon Johnson, and David Autor, published a direct challenge to the historical optimism this week. Their argument is that “pure automation technologies” do the opposite of collaborating with workers. They “commodify human expertise, rendering it less valuable and potentially superfluous.” If they are right, the institutional adaptations that eventually rescued workers in previous waves may not have the same leverage this time.

The distinction Morgan Stanley draws between augmentation and substitution is real and matters for how you read the current data. Morgan Stanley Research Economist Diego Anzoategui says: “The same technology that automates tasks can also augment workers, increase productivity and boost demand in AI-exposed sectors. So far, the data suggest early, narrow displacement, more visible among younger workers, while overall disruption remains limited.”

That framing is accurate as a description of what the aggregate data currently shows. The Acemoglu et al paper does not dispute the current data. It disputes the inference that current patterns will hold as AI capabilities accelerate.

The historical record they rely on, the same record Morgan Stanley cites, suggests that even when the long-run outcome is positive, the transition itself can immiserate two or three generations of workers before institutions catch up.

The gender dimension of this debate has received far less attention than it deserves. A research brief published in March by the International Labour Organization, drawing on data from 436 occupations across dozens of countries, established a finding that changes the shape of the entire conversation.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Female-dominated occupations are almost twice as likely to be exposed to generative AI as male-dominated ones. Around 29% of female-dominated occupations are exposed to generative AI, compared to just 16% of male-dominated occupations.

The difference is even starker at the highest automation risk: 16% of female-dominated occupations fall into the highest exposure categories, compared to only 3% of male-dominated ones.

The International Labor Organization (ILO) traces this gap to a structural feature of the labor market that predates AI entirely. Women are heavily concentrated in clerical, administrative, and business support roles, such as secretaries, receptionists, payroll clerks, and accounting assistants. Many of the tasks are routine and codifiable and therefore at higher risk of substitution by generative AI.

By contrast, men are more represented in construction, manufacturing, and manual trades, where tasks are less easily automated.

The problem compounds at the other end. The jobs being created by AI adoption, the ones carrying wage premiums and growing demand, are concentrated in engineering, cloud architecture, and AI development.

Globally, women accounted for only about 30% of the AI workforce in 2022, only 4 percentage points higher than in 2016. The displacement risk is concentrated where women work. The opportunity is concentrated where women are underrepresented.

That structural mismatch does not resolve itself through retraining programs: of the workers most at risk of losing their jobs due to AI, more than 6 million would likely struggle to cope because they are older, have limited savings, and face other barriers.

Most of those workers are in clerical and administrative jobs, roles that have historically been dominated by women.

The ILO’s senior economist Janine Berg put the core policy question plainly: The impact of generative AI on women’s jobs is not predetermined. “With the right policies, social dialogue and gender-responsive design, we can avert reinforcing existing discrimination,” she said.

If the policy choices are not made, the default outcome is an AI transition that widens the existing gender wage and employment gap.

The Morgan Stanley historical analysis and the Acemoglu et al warning are not actually incompatible once the timeline is named honestly. They operate at different scales.

Morgan Stanley’s data describes what aggregate labor markets are showing right now, and what historical patterns suggest about century-scale outcomes.

Acemoglu et al are describing the mechanism by which this wave could compress or extend the brutal middle period, the part where workers absorb the cost while waiting for institutions to catch up.

Both can be true simultaneously: the long-run historical record is broadly positive, and the transition itself can take three generations and require sustained political struggle to make those gains reach ordinary workers.

The American Enterprise Institute frames the burden of proof as belonging to what it calls “the discontinuity camp,” whose case “rests more on what AI could theoretically do than on what labor markets are actually showing.”

That framing is fair given the current aggregate data. The ILO gender numbers suggest, however, that for specific populations of workers, the discontinuity may already be arriving. Whether the institutional adaptations Morgan Stanley’s Gapen names actually happen in time, and for whom, is the political question that historical optimism cannot answer on its own.

A compilation of the Substack articles examining how the invasion already happened. You just weren’t invited. $9.95 flat fee for the bundle (PDF, ePUB), no subscription required. 2-hr reading time.

The first development is a legal one; the second, a technology one.

In the first instance, plaintiffs in the landmark case Mobley v. Workday, Inc. filed an amended complaint in late March 2026. The complaint revived California state claims and a physical disability discrimination count after a federal judge’s ruling kept core federal age discrimination claims alive.

The case involves Derek Mobley, an African-American man over 40 with anxiety and depression. He applied for more than 100 jobs through companies using Workday’s AI-powered applicant screening tools. He was rejected every time, sometimes within minutes of applying.

A federal court has allowed discrimination claims against Workday’s AI screening tools to proceed on an unprecedented “agent” liability theory. The case is advancing primarily on age discrimination under the Age Discrimination in Employment Act. The outcome of the case could impact the lives of hundreds of millions of job applicants.

The case is the first major test of whether AI vendors — not just employers — can be held directly liable for algorithmic hiring bias under federal anti-discrimination law.

The legal significance here is structural, not just factual. AI hiring tools are used by thousands of companies as a cost-saving measure.

87% of companies are now using AI-driven tools to save time and money in the hiring process. When those tools discriminate, they do so at scale, across every employer using the same platform simultaneously.

The Workday case tests a theory that has major consequences for anyone in the job market over age 40: that the vendor of the discriminating tool shares legal liability with the employer.

If the court upholds that theory, it opens class-action exposure for every major HR software company operating AI screening at scale.

A University of Washington study found that recruiters using biased AI tools mirrored the tool’s inequitable choices up to 90% of the time. The study illustrates that both the AI vendor and the employer carry liability.

If your company runs AI applicant tracking systems without documented bias testing, adverse impact analysis, and a human-in-the-loop override, the Board is sitting on a legal exposure it did not budget for.

The second development connects to the first at the level of capability.

OpenAI’s GPT-5.4, released in early March, achieves a new state of the art on GDPval. GDPval is a benchmark that tests AI agents on well-specified knowledge work across 44 occupations.

GDPval measures real work products: sales presentations, accounting spreadsheets, urgent care schedules, manufacturing diagrams. It is not a test of abstract reasoning. It is a test of the actual output that knowledge workers produce on a given Tuesday.

OpenAI’s 5.4 model matches or exceeds industry professionals in 83% of comparisons. That’s up from 70.9% for its predecessor, GPT-5.2.

On an internal benchmark simulating tasks that a junior investment banking analyst might perform, GPT-5.4 scored 87.3%, up from 68.4% for GPT-5.2. Tasks include spreadsheet modeling, scenario analysis, and data extraction.

The improvement represents a 28-point gain in under four months. The rapid improvement suggests accuracy could reach human professional-levels in the high 80s or low 90s within the next year if the trend holds.

So, AI tools are now being used to screen job applicants, manage HR workflows, and conduct tasks previously performed by junior knowledge workers. The Workday case reveals that those tools carry discrimination risks that no one has adequately audited.

The GPT-5.4 benchmark results reveal that those tools are improving fast enough that the category of “tasks requiring human judgment” is shrinking.

AI is not merely replacing tasks, but deconstructing occupations into automated and human-centric components.

“Deconstructing occupations” describes something more precise than replacement. A job that used to require a human is being split into its component tasks. Some tasks go to AI.

The remaining tasks stay with humans; those activities that require judgment, relationship management, or physical presence. The human does fewer things. Their employer employs fewer people.

In the Gulf region, financial institutions are deploying autonomous agents to handle 80% of routine compliance and reporting. The agents enable human staff to focus on high-stakes advisory and regional strategy.

The most notable change in 2026 is the move toward agentic workflows. Modern AI agents can now coordinate across departments. They can manage complex projects with minimal human oversight.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

According to PwC, roughly 44% of core worker skills have been disrupted in the last 24 months. The new gold standard for the global workforce includes:

* strategic prompting and orchestration

* the ability to manage multiple AI agents simultaneously

* AI ethics and bias mitigation

* and complex emotional intelligence , or EQ

So, the Workday case is working through a legal system designed for a world where a human manager made hiring decisions. The AI tools now making those decisions at scale were designed by engineers optimizing for efficiency.

The discrimination they produce is not intentional. That makes it harder to detect, harder to litigate, and harder to remedy.

Colorado’s AI law, set to take effect at the end of June 2026, will require employers in the state that deploy high-risk AI systems to take “reasonable care” to protect consumers from discrimination.

HR software that screens candidates, scores performance, or ranks employees is classified as high-risk AI under Colorado’s law. Colorado is the first state to impose that standard.

Whether other states follow suit before the tools spread further is an open question. The outcome has real consequences for every worker who will be filtered by an algorithm in the next job search.

A compilation of the Substack articles examining how the invasion already happened. You just weren’t invited. $9.95 flat fee for the bundle (PDF, ePUB), no subscription required. 2-hr reading time.

An analysis of regulatory filings shows that Indian IT and business process outsourcing firms may have laid off more people in the U.S. in the first three months of 2026 than in all of 2025.

WARN notices, required by U.S. law for large layoffs, have been filed by Infosys, HCL Technologies, and Hinduja Global Services across Florida, Texas, and Pennsylvania. WARN notices are the formal legal mechanism companies use to notify state governments of mass layoffs 60 days in advance.

The spike in filings is a documented, verifiable acceleration, not a trend based on projections. In 2024, Indian IT firms filed four WARN notices. In 2025, only Genpact filed one. But in the first quarter of 2026 alone, multiple notices have already been filed.

This is a red flag for United States AI-related job cuts: Indian nationals likely do not show up in layoff statistics, as Indian IT firms may have hired them as contractors and not employees. Laid off employees can claim unemployment benefits, which appear in state and federal registers; contractors have to troll for their next gig.

Indian IT companies are cutting more jobs in the U.S. mainly because AI is making some roles less necessary and big tech deals are slowing down. Most of the layoffs hit on-site workers tied to huge transformation projects.

The old model worked like this: a major U.S. bank would sign a multi-billion dollar, multi-year outsourcing contract. The Indian IT firm would deploy hundreds of engineers on U.S. soil to manage the project.

AI is shrinking the headcount those projects require, though. Productivity gains of up to 30% mean fewer people are needed for the same work. The industry is shifting from execution to collaboration, and the old model of deploying thousands of engineers onsite is fading.

The human consequence of this shift is not simply that Indian workers lose U.S.-based jobs. While not impossible, career transitions are difficult and require time — a precious commodity in any labor market.

A mid-level programmer whose role is automated may not easily transition to becoming an AI architect or a strategic advisor. The reskilling challenge is enormous. While companies are investing billions in training programs, the question is whether those programs can scale fast enough to meet the needs of displaced workers.

Business Process Outsourcing — or BPO — presents the same problem at larger scale. India employs roughly 1.65 million workers in voice support, data processing, and administrative BPO roles.

Startups like Bengaluru-based LimeChat are deploying AI agents that can handle up to 95% of customer queries without human assistance. “Once you hire a LimeChat agent, you never have to hire again,” said Nikhil Gupta, co-founder of LimeChat.

For the businesses buying this technology, the appeal is lower costs and higher scalability. The business opportunity foreshadows lower headcounts for the 1.65 million workers whose livelihoods depend on those call center roles.

Entry-level BPO roles, the foundation of the offshore talent pyramid, sit squarely in the line of fire, mirroring broader warnings that AI could dramatically shrink India’s BPO workforce by 2030.

The question now being asked across the sector is whether that displacement accelerates the transition toward higher-value knowledge process outsourcing, or KPO. KPO refers to higher-complexity work like legal research, financial analysis, and medical coding. AI already handles routine data entry and customer service calls more cheaply.

Stilll, human judgment still commands a premium, even if AI hollows out the talent pipeline that feeds it.

The question is whether enough KPO work exists to absorb the workers displaced from the lower tier.

The picture from South Asia is not entirely negative, which provides indicators for the United States labor market. In South Asia, AI-related job postings have increased, offering wages roughly 30% higher than comparable white-collar jobs.

The premium for AI skills is real and documented. The problem, however, is the gap between who is losing routine BPO work and who is positioned to capture AI-adjacent roles. Those are largely different people with different educational backgrounds and different geographic locations within India.

LinkedIn’s 2026 Labor Market Report, released in January and drawing on data from over 1.3 billion users globally, offers a frame for understanding the new job categories being created.

AI has created demand at scale, including more than 600,000 new AI-enabled data center jobs and 1.3 million new roles.

Some of the new roles include: AI engineers, forward-deployed engineers, and data annotators. AI Engineer is one of the fastest-growing jobs on LinkedIn over the past three years.

This mix of uneven hiring and AI-driven job creation marks what LinkedIn among others are calling the new-collar era: a workforce that blends knowledge work, advanced technical skills, and distinctly human strengths.

Take, for instance, the forward-deployed engineer role. It did not exist as a job title five years ago. The function is to take AI systems that work in a controlled technical environment and make them actually function inside a specific company’s existing workflows, systems, and culture.

Future Forwarded is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Building AI turns out to be considerably easier than deploying it successfully. The demand for people who can bridge that gap is real.

So, AI is creating a class of premium new roles while simultaneously eliminating a much larger class of routine roles.

The new roles pay more. They require different skills. They are concentrated in different geographies.

The consensus emerging from buyers and analysts is that what remains after AI automation is precisely where specialist human partners add the most defensible value. We’re talking , complex judgment calls, sensitive customer interactions, and regulated processes,

The question arises, though: how many of the 1.65 million displaced BPO workers, or the thousands of Indian IT engineers losing on-site U.S. contracts can retool before their windows of opportunity close.

A compilation of the Substack articles examining how the invasion already happened. You just weren’t invited. $9.95 flat fee for the bundle (PDF, ePUB), no subscription required. 2-hr reading time.