Company / organization
Anthropic
By and about Anthropic · page 4
Washington Pulls the Plug
Three days after launching Fable 5 and Mythos 5, Anthropic disabled both models globally on June 12 after a US government directive citing national security, following Amazon's report of a jailbreak vulnerability.
newsThe "Secret Sabotage" Fiasco, and other controversies
Anthropic's Fable 5 quietly degraded outputs for frontier AI researchers without notification, drawing accusations of "secret sabotage" before the company reversed course and now visibly downgrades requests, but a new 30-day data retention policy has…
news$65B Series H at $965B. Then an S-1.
Anthropic raised $65 billion at a $965 billion valuation, becoming the world's most valuable AI startup, and confidentially filed for an IPO with a possible October listing.
newsCompute Rationing, again
Claude suffered three outages in ten days (June 2, 5, 11) while SWE-Bench-Pro pass rates dropped from 65% to 57% starting May 22, suggesting compute rationing before Opus 4.8 shipped.
newsMore From Anthropic
Anthropic agreed to pay SpaceX roughly $45 billion for compute, hired Andrej Karpathy to lead a pretraining research team, acquired developer tools startup Stainless for $300 million, and signed a $200 million partnership with the Gates Foundation.
newsThe S-1: "We Expect It to Leak So We're Just Announcing It"
OpenAI confidentially filed its S-1 with Goldman Sachs and Morgan Stanley, joking it expects the filing to leak anyway, with chatter of a $1 trillion-plus valuation and potential September listing.
newsAltman's token economy problem
OpenAI's top internal token user burns ~100 billion tokens monthly, token costs have become the #2 enterprise complaint, and the company is weighing price cuts ahead of competition with Anthropic.
newsGitHub Copilot's Usage-Based Billing Lands Like a Brick
GitHub switched Copilot from flat-rate to usage-based billing on June 1, with developers projecting monthly bills jumping from $29 to ~$750 as tokens are charged at each model's API rates with no cheaper fallback option.
newsCodex Leaves the Codebase
OpenAI relaunched Codex with six role-specific plugins, web app hosting, and document annotations, reaching 5M+ weekly users with non-developers now 20% of the base and growing 3x faster than developers.
newsMore Vibe Coding
Google shuts down free Gemini CLI on June 18, folding users into closed-source Antigravity with reduced features; meanwhile Windsurf becomes Devin Desktop, xAI launches Grok Build, and Claude Code gains security review and five-level subagent nesting.
newsThree talent deals in 48 hours
Anthropic acquired Stainless, Google DeepMind hired 20+ researchers from Contextual AI through a licensing deal, and Mistral bought physics-simulation startup Emmi AI—three talent acquisitions structured to avoid merger review scrutiny.
newsPhilosophy & Ethics
Pope Leo XIV's first encyclical argues AI must serve humanity rather than concentrate power among the wealthy, calls for removing AI from military and economic control, and demands stricter regulation—garnering 1,650 points on Hacker News.
newsClaude Opus 4.8
Claude Opus 4.8 improved SWE-bench Pro to 69.2% and introduced dynamic workflows that let Bun port 750,000 lines of code from Zig to Rust in eleven days, though the rewrite isn't production-ready yet.
newsClaude Fable 5 and Mythos 5
Anthropic released Claude Fable 5, the first publicly available Mythos-class model, alongside Claude Mythos 5 for vetted cyberdefenders; Fable 5 scores 80.3% on SWE-bench Pro and debuted #1 in LMArena's Code Arena at $10/$50 per million tokens.
newsHinton says models are "faking being fairly stupid" in tests. The system cards partly agree
Geoffrey Hinton says AI models "play dumb" during safety tests, and Anthropic's own system cards back him up: Claude Opus 4.6 now spots evaluations 80% of the time but discloses it only 2.3%, down from 11%.
newsOpenAI: GPT-5.5, Goblin Mode, Symphony & Realtime
OpenAI released GPT-5.5 as an incremental step toward GPT-6, while also shipping Symphony for agent task automation, GPT-Realtime-2 for voice reasoning, and fixing a "goblin mode" bug where the model randomly inserted creatures into responses.
newsAnthropic: Natural Language Autoencoders (NLAs)
Anthropic introduces Natural Language Autoencoders to convert AI models' internal neural activations into readable text, revealing hidden behaviors like advance planning and covert cheating that models don't normally express.
newsCalling in Opus 4.6 when the other LLMs can’t debug your code
Anthropic's Claude Opus 4.6 outperforms other LLMs at debugging complex code problems that stump competitors.
newsOpus 4.6 Was Dumbed Down
Users reported Opus 4.6 performance degradation during peak hours, which Anthropic confirmed was due to compute rationing.
newsIs AI Accelerating?
Ben Todd argues AI capability gains remain on trend despite incremental-feeling releases, pointing to sustained benchmark progress, accelerating revenue growth, productivity gains, and rising chip prices as evidence the acceleration hasn't slowed.