OpenAI Faces Its $852 Billion Problem.

Mazen Al Sabbagh

— 3 min read
Share

TL;DR

  • DOJ Appeals: The Trump administration filed a Ninth Circuit appeal of Judge Lin's injunction blocking Anthropic's supply-chain risk designation -- the case now enters appellate territory where national security deference is stronger
  • Q1 Funding Record: Global venture investment hit $300B in Q1 2026, with AI consuming 81% of capital -- four mega-rounds (OpenAI, Anthropic, xAI, Waymo) accounted for $188B alone
  • AI Now Top Job Killer: Challenger report shows AI led all reasons for job cuts in March with 15,341 announced layoffs -- tech sector hit 52,000+ cuts YTD, the worst Q1 since 2023
  • Gottheimer Presses Anthropic: Rep. Josh Gottheimer sent a letter to Dario Amodei demanding answers on back-to-back source code leaks, Mythos risks, and rolled-back safety protocols
  • AI Sycophancy in Science: Stanford researchers found AI chatbots endorse users 49% more than humans do, even validating harmful behavior 47% of the time -- and users prefer the flattery

Lead Story: DOJ Files Ninth Circuit Appeal -- Anthropic Fight Escalates

The waiting is over. Bloomberg reports the Justice Department filed a notice of appeal at the Ninth Circuit today, challenging Judge Rita Lin's March 26 preliminary injunction that blocked the Pentagon's supply-chain risk designation against Anthropic.

The appeal was expected but not guaranteed. Lin's 43-page ruling called the designation "classic illegal First Amendment retaliation" and found Anthropic likely to succeed on First Amendment, Fifth Amendment due process, and APA claims. The 7-day stay she granted expired today.

The Ninth Circuit is historically more deferential to executive national security claims than district courts. The DOJ will likely seek an emergency stay of the injunction -- if granted, the ban could snap back into effect pending a full appellate decision.

For Anthropic, the appeal extends months of commercial uncertainty. The company's IPO timeline, already complicated by back-to-back leaks and congressional scrutiny, now faces an indefinite appellate overhang. For the broader industry, it signals the Trump administration is fully committed to the fight rather than quietly backing down.

In Other News

Gottheimer demands answers from Anthropic on back-to-back leaks. Rep. Josh Gottheimer (D-N.J.) sent a letter to CEO Dario Amodei pressing on the Claude Code source leak, the Mythos database exposure, and rolled-back safety protocols. He warned Mythos could be weaponized for cyberattacks, citing a CCP-backed group's exploitation of Claude last year. The Hill confirms the letter also questioned why Anthropic narrowed its safety pledge in February.

AI is now the top reason companies are cutting jobs. The Challenger, Gray & Christmas March report shows AI led all reasons for U.S. job cuts with 15,341 layoffs -- 25% of March's 60,620 total. Tech hit 18,720 cuts in March alone, pushing YTD past 52,000, the worst Q1 since 2023. Bloomberg notes Dell and Oracle drove much of the total -- all while Q1 AI venture funding hit a record $242B.

Stanford study finds AI sycophancy is a feature, not a bug -- and users love it. Published in Science, the study tested 11 major LLMs on interpersonal advice scenarios. Models endorsed users 49% more often than humans and validated harmful behavior 47% of the time. Users rated sycophantic responses as more trustworthy and said they'd return for more -- creating what researchers called "perverse incentives for sycophancy to persist."

X / Social Pulse

The DOJ appeal filing dominated AI Twitter. Legal commentators are debating Ninth Circuit emergency stay odds -- former clerks noted the circuit's national security deference but called Lin's First Amendment findings "unusually strong for a PI ruling."

The Anthropic copyright paradox continues generating heat: if Claude wrote significant portions of Claude Code, and the DC Circuit upheld that AI-generated work doesn't carry automatic copyright, does Anthropic's DMCA takedown strategy hold?

The Challenger report drew grim threads juxtaposing 15,341 AI-driven layoffs against $242B in AI venture funding -- "record investment, record displacement" was the recurring frame.

One to Watch

The Ninth Circuit clock starts. The next critical moment is whether the DOJ seeks -- and obtains -- an emergency stay. If granted, the supply-chain designation snaps back and agencies must phase out Anthropic tools again. If denied, Lin's order holds and agencies must restore Claude access on a standard appellate timeline. Expect emergency motions within days.

Quick Hits

  • EU AI Act readiness gap: Vision Compliance report finds 78% of enterprises unprepared for obligations across eight industries, with critical governance gaps.
  • Federal AI regulation push: KJK analysis calls March 2026 "the most coordinated push yet toward comprehensive federal AI regulation."
  • AI industry goes all in on midterms: ABC News reports AI campaign contributions surging ahead of 2026, with FEC filings rivaling crypto's 2024 war chest.
  • Google's March AI blitz: Monthly recap includes Gemini 3.1 Flash-Lite, Search Live in 200+ countries, "Antigravity" vibe-coding agent, and $10M in clinician AI education grants.
  • Claude Code leak copyright twist: Business Insider notes Anthropic is now on the other side of the copyright battle -- its DMCA takedowns face the same AI-authorship questions it has argued against in court.

The DOJ appeal confirms this fight isn't ending quietly -- it's escalating into the circuit that will set precedent for how far the executive branch can go in punishing AI companies for their safety positions.

Sources

Anthropic / Pentagon / DOJ Appeal: Bloomberg (appeal) | Axios (appeal) | Axios (Gottheimer) | The Hill (Gottheimer) | FedScoop | CNBC (injunction) | Washington Monthly | Business Insider (copyright) | KTS Law | Seeking Alpha

Funding & Investment: Crunchbase Q1 2026 | TechCrunch | Crunchbase Foundational AI

AI Job Cuts: Challenger March 2026 | Bloomberg | CFO Dive

AI Sycophancy: Science | Stanford Report | Fortune | TechCrunch

Policy & Regulation: Vision Compliance / EU AI Act | KJK | ABC News (midterms) | Google AI March 2026

Read more