Three federal agencies produced AI adoption rates of 18%, 41% and 78% for the same quarter; the Bureau of Labor Statistics (BLS) skipped its scheduled GenAI publication on 14 April and the Federal Reserve Bank of New York filled the gap with a survey showing 62% of workers expect AI-driven unemployment within twelve months. Stanford Digital Economy Lab calculates the real AI labour impact at roughly 34 times the Challenger, Gray & Christmas declared tally.
Declared AI cuts are one-thirty-fourth of the real impact, and the measurement system to prove it is being pre-empted electorally.
The Federal Reserve Board published a reconciliation paper on 3 April comparing three federal instruments that describe the same late-2025 economy as 18%, 41% or 78% AI-adopted. The gap is structural.
The Federal Reserve Board published a reconciliation paper on 3 April 2026 showing 3 separate federal surveys produced AI adoption rates of 18%, 41% and 78% for the same late-2025 period. Daily AI use sits at 12% of the US workforce; weekly use at 35.2%.
No federal agency has chosen a canonical unit of measurement. The 4.3x range is not noise; it reflects 3 incompatible survey architectures. US displacement policy has no statutory baseline to legislate against.
Erik Brynjolfsson's Stanford Digital Economy Lab applied the JOLTS hiring rate to the nonfarm workforce and found roughly a million annual hires are not happening. Against the declared AI layoff count, the ratio is 34 to 1.
Stanford Digital Economy Lab applied the February 2026 Bureau of Labor Statistics Job Openings Survey, which hit a 3.1% hiring rate (lowest since April 2020), to the 158.6 million nonfarm workforce. Result: roughly 950,000 to 1 million fewer annual hires than the 2023 pace. Against 27,645 declared AI layoffs through March, the ratio runs 34-to-1.
Workers aged 22-25 in AI-exposed occupations show a 16% employment decline since late 2022. AI displacement runs through hires that never happen.
The Bureau of Labor Statistics skipped its scheduled GenAI workplace paper on 14 April. The Federal Reserve Bank of New York published its own survey the same day, finding 62% of American workers now expect AI to raise unemployment within twelve months.
The Bureau of Labor Statistics skipped its scheduled GenAI workplace publication on 14 April. The Federal Reserve Bank of New York filled the gap, publishing Survey of Consumer Expectations data: 39% of workers use AI tools, 62% expect AI to raise unemployment, and access runs 4.2x higher for workers earning over $200,000 than those earning under $50,000.
Only 15.9% of US employers offer AI training. Workers accept an 11.4% salary cut to join a role that provides it.
Leading the Future, a super PAC backed by OpenAI president Greg Brockman, Andreessen Horowitz and Palantir co-founder Joe Lonsdale, has raised more than $125 million to defeat pro-regulation candidates in the 2026 US midterms.
A super PAC called Leading the Future, backed by OpenAI president Greg Brockman, Andreessen Horowitz and Palantir co-founder Joe Lonsdale, raised over $125 million to defeat pro-regulation candidates in the 2026 US midterms. Total AI industry midterm spending approaches $150 million.
The PAC targets primary elections and focuses on workforce disclosure bills, not operational restrictions. It prevents the evidence base that would make future regulation viable from forming. GZERO polling puts 63% of Americans expecting AI to reduce employment.
The UK AI Security Institute's independent evaluation of Claude Mythos Preview found no single-task superiority over rival models, but confirmed a genuine autonomous capability: a 32-step attack chain equivalent to 20 hours of trained-human work.
The UK AI Security Institute published its evaluation of Claude Mythos Preview on 15 April. On isolated tasks Mythos scored above 85%, within 5-10 points of GPT-5.4 and Codex 5.3. In the 32-step "The Last Ones" benchmark, Mythos completed a sequence the Institute estimates at 20 hours of trained-human work, autonomously.
That 20-hour figure is the first external confirmation of Mythos's attack-chaining capability , the dimension that most directly substitutes for salaried knowledge-work at the task level.
The Islamic Revolutionary Guard Corps named Stargate UAE, the $500 billion OpenAI, SoftBank and Oracle joint venture, in a 1 April military targeting video. Amazon Web Services subsequently declared hard-down status for multiple zones after Iranian strikes on Bahrain and Dubai.
Iran's Islamic Revolutionary Guard Corps named Stargate UAE, the $500 billion OpenAI-SoftBank-Oracle joint venture, as a military target in a 1 April video. Amazon Web Services declared hard-down status across multiple Gulf zones after Iranian strikes on Bahrain and Dubai. Nvidia and Apple were named in the same video.
Oracle's $156 billion data centre programme, partly funded by terminating up to 30,000 workers, now includes physical assets in an active war zone, uninsurable under standard commercial policies.
Anthropic's 244-page Alignment Risk Update for Claude Mythos Preview abandoned the AI Safety Level capability threshold framework for autonomy-focused threat models, and added Broadcom, CrowdStrike, NVIDIA, Palo Alto Networks and Cisco to the Glasswing partner list.
United StatesAnthropic published a 244-page safety update for Claude Mythos Preview on 7 April, abandoning its AI Safety Level framework for autonomy-focused risk models. Project Glasswing expanded to add Broadcom, CrowdStrike, Nvidia, Palo Alto Networks and Cisco. Over 99% of the vulnerabilities Mythos found remain unpatched.
Every Glasswing partner must rebuild its internal risk framework mid-deployment, while sitting on a shared list of unpatched flaws in software all 17 partners depend on daily.
The Bank of England Financial Policy Committee asked the BoE and the Financial Conduct Authority for further work on agentic AI in payments and financial markets, noting that three-quarters of UK financial firms now deploy AI.
The Bank of England's Financial Policy Committee, in its April 2026 record, directed the Bank and the Financial Conduct Authority to do further work on agentic AI in payments and markets. The FPC noted that 75% of UK financial firms deploy AI and assessed that systemic risk is “likely to increase rapidly”.
Agentic AI executes multi-step decisions without human confirmation. A correlated failure across 75% of UK financial firms could propagate through payments faster than any existing circuit-breaker can respond.
The Office for National Statistics reported UK vacancies unchanged at 721,000 for a sixth consecutive publication, payrolled employment down 96,000 year-on-year, and real wage growth at 0.4%.
The UK Office for National Statistics recorded vacancies unchanged at 721,000 for a 6th consecutive publication in March 2026. Payrolled employees fell 96,000 year-on-year. Real wage growth registered at 0.4% against 3.8% nominal, with the ONS collecting no AI-specific labour market breakdown.
A static vacancy count alongside falling payrolled employment points to structural stasis: firms letting attrition run without replacement. Britain’s statistical agency is not measuring the mechanism the Bank of England has now classified as a systemic financial risk.
The EU Digital Omnibus second trilogue is scheduled for 28 April. The employer AI literacy obligation stripped by Parliament on 26 March remains contested entering negotiations.
The EU Digital Omnibus second trilogue is scheduled for 28 April 2026. Employer AI literacy obligations, requiring firms to ensure workers understand how AI systems are deployed on them, were stripped by the European Parliament on 26 March. The high-risk employment deadline is fixed at 2 December 2027.
If the literacy clause survives, EU workers gain a statutory right to understand AI used to manage or evaluate them. If dropped, that right disappears from this legislative cycle until December 2027.
Oracle's Massachusetts WARN Act filing remains absent as of 13 April. The 60-day clock from the 31 March cuts expires around 30 May; Burlington offices remain unrepresented in any state disclosure.
Oracle's Massachusetts WARN Act filing remained absent as of 13 April, despite the 60-day clock from late-March terminations expiring around 30 May. Prior Oracle filings covered Washington state (491 positions) and Missouri (539), fewer than 4% of up to 30,000 affected workers.
Oracle chose not to file in Massachusetts: it filed where enforcement risk was unambiguous and skipped where the 100-worker site threshold was contestable. Massachusetts tests whether WARN Act law produces any consequence for AI-driven workforce reductions.
JPMorgan Chase CEO Jamie Dimon confirmed at the bank's February investor meeting that the bank has 'displaced people from AI' and offers them other jobs. The bank has committed $600 million annually to retraining and tied engineer performance reviews to AI tool adoption across 65,000 staff.
JPMorgan Chase CEO Jamie Dimon confirmed at the bank's February 2026 investor meeting that JPMorgan has “displaced people from AI” and offers them other jobs. The bank committed $600 million annually to retraining and tied engineer reviews to AI adoption across 65,000 staff.
Research firm MetaIntro found only 6% of companies globally are reskilling workers for AI. JPMorgan holds 1 of 12 original Glasswing partner seats giving privileged access to Anthropic's Mythos model, making the $600 million a statistical outlier.
Escalating on three axes simultaneously. Politically, $150 million in PAC spending is targeting the measurement caucus before the 2026 midterms, compressing the legislative window for statutory disclosure requirements. Militarily, the IRGC's explicit naming of Gulf AI infrastructure as a target introduces uninsurable war risk onto the balance sheets of Oracle, AWS and OpenAI. Institutionally, the abandonment of AI Safety Level thresholds in favour of autonomy-focused risk models removes the externally verifiable compliance architecture at the moment when attack-chaining capability has been independently confirmed by a state evaluator.
