Anthropic just launched Claude Fable 5 — a public release of the same underlying model as Mythos 5, held back from public access until today by a classifier layer that redirects sensitive queries to the older Opus 4.8 rather than refusing them outright. Last week, an Anthropic blog post said the safety measures needed for a broad Mythos-class release had "not yet" been developed; today the company says they're good enough to ship, at double Opus 4.8's output price. The gap between those two statements is the story.

Sam Altman, Dario Amodei, Demis Hassabis, and Mustafa Suleyman co-signed a letter to Congress this week asking for mandatory screening on synthetic DNA orders — citing evidence that AI systems now outperform PhD-level virologists on lab procedure questions. What the letter frames as a fresh bipartisan opportunity has a quieter backstory: the Biden administration's nucleic acid screening framework was ordered dismantled by Trump executive order in May 2025, a 90-day replacement deadline passed months ago with nothing published, and the Senate bill the letter is designed to move has been sitting in committee since early 2026. The ask is a supply-chain fix, not an AI fix — and the gap between what these companies are asking Congress to build and what their own researchers have shown can already defeat existing screening is the tension worth watching.

I'm ready to write the BotCity episode description. Please paste the research brief and I'll get to work.

Anthropic just dropped a new flagship model 41 days after the last one, which is not a cadence, it is a correction, and the gap between those two sentences is the whole story. Opus 4.8 genuinely leads the field on SWE-Bench Pro at 69.2% versus GPT-5.5's 58.6%, and the codebase migration claim is the most operationally significant thing any AI lab has shipped this year if it holds in production. But the company is also sitting on its most advanced model over unspecified cybersecurity concerns, has not explained what broke in Opus 4.7, and is now asking enterprise teams to run hundreds of parallel subagents on production code, so the benchmark table is not the whole question here.

Anthropic just shipped a model that writes a text-to-speech engine in Rust, generates audio, runs it through a speech recognizer, and verifies its own output with zero humans in the loop — and barely anyone noticed because the headline was the 900 billion dollar valuation. Meanwhile, they have a stronger model called Mythos scoring 83.1 percent on a cybersecurity benchmark that Opus 4.7 only hits 73.1 on, and they are quietly keeping it locked away from the public while negotiating a 50 billion dollar funding round and preparing for an IPO. The self-verifying AI shipped two weeks ago and the company sitting on an even more powerful one while structuring its own public offering is not confused about what it is doing.

A nonprofit posted a YouTube video about a Georgia data center's noise and water use, no threats, no calls to action, and it ended up in federal intelligence files next to murder investigations. WIRED obtained over a thousand pages of unpublished DHS, FBI, and fusion center reports showing a surveillance category called "anti-tech violent extremism" that has no published definition, no published threshold, and is being applied to people who show up to budget meetings. The firebombings were real, the murder charges are real, but so is the fact that a municipal treasurer in Michigan just resigned in tears over data center threats while the NYPD is formally classifying concern about AI safety as a precursor ideology to extremism.

Anthropic's AI model Claude Mythos Preview just cracked a 27-year-old bug in OpenBSD and broke Apple's brand-new M5 hardware security feature in five days flat — the one Apple spent five years and billions building. Over 99% of what Mythos has found so far is still unpatched, and that number is not a glitch in the data, that is the entire story. The patching backlog was already a disaster before this model existed, and now it has a power tool feeding it around the clock.

OpenAI's CFO said the company was not ready to go public in 2026, and seven days later, multiple outlets reported it is preparing to file for an IPO anyway. In the window between those two events, a headline surfaced about OpenAI making billions by promising future purchases from suppliers, which is exactly the kind of move that makes a balance sheet look healthier right before investors see it. Someone in that building decided the IPO window mattered more than the readiness, and the calendar got moved to match a decision that was probably already made.

Google built a video AI that can already generate long-form clips and then deliberately shipped it capped at 10 seconds, not because the model can't do more, but because the pricing, compute costs, and liability exposure around longer video aren't figured out yet. The "most users won't want longer videos" explanation sounds like product strategy, but it's really a placeholder while Google works out what this thing actually costs to run at scale and who's legally responsible when a verified user deepfakes someone. The real story from Google I/O isn't the capability, it's that the capability race everyone's covering is just the warmup, and the actual competition starts when the cap lifts.

Elon Musk's entire $38 million was gone by 2020, spent before he even had legal standing to sue, and the deadline to file ran out in August 2021 — he filed in 2024. His own board rep signed off on the exact deal structure he says blindsided him, his own advisers can't name a single condition he claims was violated, and the judge has already said if the jury finds the claims are time-barred, the whole case is over. The most expensive AI courtroom drama in history might not end with a verdict on betrayal or mission or whether Sam Altman lies — it might end because Elon Musk missed a deadline.