# Starlight AI Daily — February 25, 2026 *Tracking the intersection of artificial intelligence, policy, and global markets* --- ## The Week That Shook AI *A convergence of forces—Pentagon ultimatums, corporate policy rollbacks, and market panic—suggests the AI industry's moment of truth has arrived.* --- ## The Pentagon's Friday Deadline On Tuesday, February 25, 2026, Defense Secretary Pete Hegseth met with Anthropic CEO Dario Amodei at the Pentagon for what was described as a ["tense" face-to-face conversation](https://www.techbrew.com/stories/2026/02/25/pentagon-deadline-anthropic). The outcome: a deadline. By Friday at 5:01 PM, Anthropic must agree to allow its Claude AI models to be used for what the Pentagon is calling "unrestricted military applications"—specifically, domestic surveillance and fully autonomous weapons systems. > "The next couple days might show just how resilient AI ethics pledges are in the face of potential business losses and DC pariah status." > — [Tech Brew](https://www.techbrew.com/stories/2026/02/25/pentagon-deadline-anthropic) This is not a minor negotiation. The Pentagon has made clear it will invoke the [Defense Production Act](https://www.dw.com/en/us-pentagon-gives-ultimatum-to-anthropic-over-ai-curbs-report/a-76111915) to force compliance if necessary, and has threatened to declare Anthropic a "supply chain risk"—a designation that would effectively blacklist the company from federal contracting. At stake is a [$200 million contract](https://thehill.com/policy/technology/5754539-anthropic-ai-pentagon-dispute/), and more importantly, the precedent for how AI companies interact with military and intelligence agencies. ### Key Developments - **Deadline**: Friday, February 27, 2026 at 5:01 PM - **Stakes**: $200 million contract, designation as "supply chain risk" - **Anthropic's Position**: Refuses to allow use for domestic surveillance or fully autonomous weapons Anthropic's Claude was, until recently, the [only AI model approved](https://www.techbrew.com/stories/2026/02/25/pentagon-deadline-anthropic) for use in classified U.S. government operations. --- ## The Quiet Retreat from Safety Pledges While the Pentagon was delivering its ultimatum, Anthropic was executing a different kind of rollback—one that received far less attention but may prove more consequential in the long run. On February 25, 2026, Anthropic released an updated version of its [Responsible Scaling Policy](https://www.anthropic.com/responsible-scaling-policy). The changes were significant: ### Policy Changes - **Removed**: Previous commitment to halt development if safety procedures couldn't keep pace - **Replaced with**: "Nonbinding but publicly-declared" goals > "The AI industry has not reached a consensus on risks as it had previously hoped... pausing development while others forge ahead would result in a world that is less safe—the developers with the weakest protections would set the pace." > — [Anthropic Responsible Scaling Policy Update](https://www.anthropic.com/responsible-scaling-policy) The timing raises questions. The policy update came on the same day as the Pentagon meeting, leading some analysts to wonder whether [the company was already preparing to lower its guardrails](https://thehill.com/policy/technology/5754539-anthropic-ai-pentagon-dispute/) before the meeting took place. --- ## The Citrini Panic On Monday, February 24, U.S. stocks slid sharply after a research memo from Citrini Research went viral across trading desks. ### Report Overview - **Title**: ["The 2028 Global Intelligence Crisis"](https://www.citriniresearch.com/p/2028gic) - **Published**: February 22, 2026 - **Authors**: Citrini Research, Alap Shah (Lotus Technology Management) - **Framing**: Scenario, not prediction ### Key Projections | Metric | Projection | |--------|------------| | Unemployment | 10%+ by mid-2028 | | S&P 500 | 38% decline from October 2026 highs | | White-collar cuts | 5% within 18 months | > "If AI automates high-earning knowledge workers—software engineers, financial analysts, lawyers—at the rate that current trajectory suggests is possible, the consumption engine that drives the economy sputters." > — [FXStreet Analysis](https://www.fxstreet.com/analysis/the-citrini-report-how-a-debatable-ai-narrative-can-shake-wall-street-202602241415) The memo caused a [mid-afternoon London selloff](https://www.fxstreet.com/analysis/the-citrini-report-how-a-debatable-ai-narrative-can-shake-wall-street-202602241415), demonstrating how fragile AI optimism has become. --- ## The LinkedIn Reality Check Against these dramatic headlines, labor market data offers a more nuanced picture. ### Job Market Statistics - **600%+**: Increase in AI mentions in U.S. job listings over past three years - **1.3 million**: New "AI-enabled jobs" globally - **600,000**: New AI-enabled data center positions created in the past year - **56%**: Professionals planning to job hunt in 2026 - **76%**: Professionals who don't feel prepared for the new job market > "Companies can grow their AI talent pipeline by 8.2x globally by focusing on skills over degrees." > — [LinkedIn Labor Market Report](https://economicgraph.linkedin.com/content/dam/me/economicgraph/en-us/PDF/linkedIn-labor-market-report-building-a-future-of-work-that-works-jan-2026.pdf) This is not (yet) a story of mass displacement. According to the [World Economic Forum](https://www.weforum.org/stories/2026/01/ai-has-already-added-1-3-million-new-jobs-according-to-linkedin-data/), AI is creating jobs faster than it's displacing them. --- ## Thread Summary ### Three Tensions Converging 1. **Government vs. Industry**: Pentagon demands unrestricted access; AI companies face pressure to comply or lose contracts 2. **Safety vs. Competition**: Companies weakening safety commitments citing competitive pressure 3. **Productivity vs. Displacement**: Markets pricing endless growth; scenario models suggest crisis potential ### What to Watch - **Friday, Feb 27**: Pentagon deadline for Anthropic response - **Coming Weeks**: Whether other AI companies follow Anthropic's policy retreat - **Q2 2026**: Labor market data will reveal whether job creation keeps pace with displacement fears --- ## Sources - [Tech Brew - "A red line deadline for Anthropic"](https://www.techbrew.com/stories/2026/02/25/pentagon-deadline-anthropic) - [Deutsche Welle - "Pentagon gives ultimatum to Anthropic over AI curbs"](https://www.dw.com/en/us-pentagon-gives-ultimatum-to-anthropic-over-ai-curbs-report/a-76111915) - [Al Jazeera - "Anthropic vs the Pentagon"](https://www.aljazeera.com/news/2026/2/25/anthropic-vs-the-pentagon-why-ai-firm-is-taking-on-trump-administration) - [The Hill - "Anthropic narrows AI safety policy pledge"](https://thehill.com/policy/technology/5754539-anthropic-ai-pentagon-dispute/) - [Citrini Research - "The 2028 Global Intelligence Crisis"](https://www.citriniresearch.com/p/2028gic) - [FXStreet - "The Citrini Report"](https://www.fxstreet.com/analysis/the-citrini-report-how-a-debatable-ai-narrative-can-shake-wall-street-202602241415) - [LinkedIn Labor Market Report](https://economicgraph.linkedin.com/content/dam/me/economicgraph/en-us/PDF/linkedIn-labor-market-report-building-a-future-of-work-that-works-jan-2026.pdf) - [World Economic Forum - "AI has already added 1.3 million new jobs"](https://www.weforum.org/stories/2026/01/ai-has-already-added-1-3-million-new-jobs-according-to-linkedin-data/) --- *Starlight AI Daily — Independent analysis of AI industry developments*