AI Intelligence Digest
TIMPS
PostCards
The Alignment Problem Isn't What You Think It Is
Issue 040 March 28, 2026 Hyderabad, India
⬡ Built with the TIMPS Ecosystem
Latest Issue
01 · LEAD
AI Safety & Governance

Federal Judge Blocks Pentagon's Anthropic Ban in Landmark First Amendment Ruling

US District Judge Rita Lin issued a blistering 43-page injunction on March 26, calling the government's designation of Anthropic as a 'supply chain risk' an 'Orwellian' act of retaliation — the first time a federal court has ruled that the US cannot punish an AI company for refusing to remove ethical safeguards.
Injunction Blocked Pentagon blacklist on First Amendment grounds

US District Judge Rita Lin in California issued a preliminary injunction blocking the Pentagon's classification of Anthropic as a 'supply chain risk,' ruling that the government had retaliated against the company for refusing to remove ethical guardrails on military AI use. 'Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary and saboteur of the U.S. for expressing disagreement with the government,' Lin wrote. The Pentagon has 7 days to seek an emergency stay from the Ninth Circuit.

The dispute originated in February when Anthropic refused Pentagon demands to modify its acceptable use policies to allow mass domestic surveillance and fully autonomous weapons. Defense Secretary Pete Hegseth responded by designating Anthropic a supply chain risk — a label previously reserved for foreign adversaries like Huawei — and ordered all federal agencies to purge Claude within 180 days. Anthropic sued on March 9, and the case has become a landmark test of whether AI companies can enforce ethical red lines in government contracts.

CNBC, LA Times, Reuters
02 · CAPITAL
Industry & Finance

OpenAI Closes $110 Billion Round at $730 Billion Valuation — and Immediately Expands It

The same week Anthropic won its court victory, OpenAI announced it had closed what was then the largest funding round in Silicon Valley history, anchored by Amazon, Nvidia, and SoftBank — and promptly kept the round open for more investors.
$110B Closed Feb 27, expanded to $122B by Apr 1

OpenAI announced the close of a $110 billion funding round at a $730 billion pre-money valuation on March 31, with Amazon committing $50 billion (including $35 billion contingent on an IPO or AGI milestone), Nvidia and SoftBank each contributing $30 billion. SoftBank separately secured a $40 billion bridge loan on March 27 to backstop its commitment, with payments structured in quarterly tranches of $10 billion. The round remained open for additional investors and would eventually reach $122 billion by early April.

At $730 billion pre-money, OpenAI's valuation exceeded every publicly listed company except Apple, Microsoft, Nvidia, and Alphabet. The company reported $2 billion in monthly revenue and over 900 million weekly active users. Amazon's $35 billion conditional tranche effectively makes a 2026 IPO a requirement of OpenAI's financing structure rather than a voluntary strategic decision.

CNBC, Bloomberg, OpenAI Blog
03 · PRODUCT
Product & Strategy

OpenAI Shuts Down Sora After Burning $15 Million Per Day on Video Generation

The video generation platform that once dazzled the world with '69 Ways of Looking at an AI' was quietly killed after internal documents revealed it was costing 7x more to run than it earned — and Disney walked away from a $1 billion partnership.
$15M/day Operating cost vs $2.1M daily revenue

OpenAI confirmed the shutdown of Sora, its AI video generation platform, after internal documents obtained by multiple outlets revealed the service was burning approximately $15 million per day in compute costs while generating only $2.1 million in daily revenue. The financial hole was compounded by Disney's decision to walk away from a $1 billion partnership deal that would have integrated Sora into the studio's production pipeline.

The Sora shutdown marks the highest-profile product failure in OpenAI's history and raises questions about the economics of generative video. OpenAI's own internal forecast reportedly projected profitability for Sora no earlier than 2031. The API will continue operating until September 24, 2026 to allow existing customers to migrate to alternatives like Google's Veo 3.1 or Kling 3.0.

Variety, Deadline, TechCrunch
04 · SIGNALS
5 Things Worth Knowing

5 Key Signals

01
Leaked Altman Slack messages show OpenAI CEO claimed he 'tried to save' Anthropic while privately seeking the Pentagon deal for himself
Axios ↗
02
Microsoft files amicus brief supporting Anthropic in Pentagon dispute — despite being OpenAI's primary investor
CNBC ↗
03
Trump appoints Zuckerberg and Huang to new PCAST advisory council; Musk and Altman excluded
The Hill ↗
04
Apple opens Siri to third-party AI services — Gemini, Claude, and others can now be default assistants in iOS 27
Apple Insider ↗
05
~200 state-level AI bills tracked by end of March as US regulatory patchwork intensifies
Manatt Health ↗
05 · THEMES
This Week in Focus

Top Themes

AI Safety Precedent
Military AI Red Lines
Video Gen Economics
Platform AI Gatekeeping
State-Level Regulation
Tool of the Week
Cursor
The AI-native IDE that surged past 2M developers as Claude Code and Codex plugins made it the most popular agentic coding environment.