AI Digest

Everything you need to know about AI

Bookmark this page. The most important things happening in AI, synthesized from dozens of sources into a two-minute read.

Updated every 48 hours Last updated June 14, 2026 · 42 sources · 191 articles · 48h window View past digests → Subscribe via RSS

  1. US government forces Anthropic to pull Claude Fable 5 and Mythos 5 offline.

    Citing national security, the Commerce Department issued an unprecedented export-control directive barring all foreign nationals from accessing the two top-tier models, so Anthropic cut off access worldwide, even for its own foreign employees. Reporting ties the move to a narrow "jailbreak" surfaced by Amazon security research and CEO Andy Jassy's talks with the White House, and Anthropic publicly disputed that a narrow flaw justifies recalling a model used by hundreds of millions.

    Sources Wired ·The Verge ·TechCrunch ·VentureBeat

  2. Meta's new AI unit is described as being in open revolt.

    Multiple reports portray the roughly 6,500-person superintelligence group as chaotic and demoralized, with staff openly rejecting a companywide AI hackathon and internal meetings turning hostile. Separately, Meta is reportedly unwinding its $2 billion Manus acquisition after Beijing ordered the deal reversed.

    Sources Wired ·TechCrunch ·TechCrunch

  3. Google sues a Chinese network that weaponized Gemini for mass scams.

    The complaint says a Telegram-based group called "Outsider Enterprise" used AI to spin up scam sites and blast roughly 2.5 million fraudulent text messages in two weeks, targeting hundreds of thousands of victims. The case is an early test of holding criminals accountable for AI-automated fraud.

    Sources Ars Technica ·TechCrunch ·The Register

  4. AI's reliability problem keeps surfacing in embarrassing ways.

    KPMG pulled a report on AI usage after most of its citations turned out to be apparently hallucinated, with GPTZero claiming only 5 of 45 citations matched their sources. Meanwhile a court ruled Google legally liable for false statements its AI Overviews generate, and Google researchers proposed "faithful uncertainty" to let models offer best guesses instead of confidently inventing answers.

    Sources TechCrunch ·The Register ·Wired ·VentureBeat

  5. The backlash against AI data centers is hardening into real friction.

    Roughly $130 billion in data-center projects have been blocked by local protests so far this year, and Amazon admitted to using 2.5 billion gallons of water to cool its facilities last year. Debate continues over how much of the opposition is organic versus blamed on foreign interference, while SK Hynix said it will triple memory output to feed AI demand.

    Sources Ars Technica ·The Register ·The Register

  6. Open-weight and non-US labs keep pressing the model race.

    Zhipu's GLM 5.2 and Moonshot's Kimi K2.7-Code both shipped this week, the latter claiming 30% leaner reasoning though practitioners questioned its benchmarks. France's Mistral is reportedly raising €3 billion at a roughly €20 billion valuation, nearly double its previous round.

    Sources Hacker News ·VentureBeat ·TechCrunch