Claude

exponential time hypothesis refuted what it means b p versus np classes diagram

Exponential Time Hypothesis Refuted? What Frontier AI Labs’ Claim Means

OpenAI and Anthropic say their models independently refuted the Exponential Time Hypothesis, a 27-year-old pillar of complexity theory. This guide explains what ETH claims, what a subexponential 3-SAT algorithm does and does not change, who is checking the proofs, and what the verdict means for algorithms, cryptography and AI-generated science.

Read more
Anthropic S-1 closing the loop shown as three Russian nesting dolls in a row on a violet hexagonal plinth

The Anthropic S-1 Filing Is Also an AI Doomsday Primer

Anthropic’s public S-1, filed on 26 October 2026, devotes 94 pages to risk factors that read like an AI doomsday primer: AI that helps build its successors, autonomous agents, self-preserving behaviour, weapons uplift, cyberattacks and robots. We quote each mechanism with page numbers, show what the filing says Anthropic has already done, and explain what it means for businesses using Claude.

Read more
Editorial illustration representing the reported upcoming Claude Fable 5.5 AI model

Claude Fable 5.5: Anthropic’s Next AI Model May Be Nearing Release

Anthropic may be preparing another major Claude release. A report published by TestingCatalog on October 17 says Claude Fable 5.5 has reportedly reached the company’s partners, with a public launch possibly close. The report draws on leaks and traces in third-party software—not an announcement from Anthropic—so the timing and specifications remain unconfirmed. The prospect of […]

Read more
prosumer ai olivia moore a16z state of consumer ai a newsstand kiosk with tiered magazine racks

a16z’s Olivia Moore on the State of Consumer AI

Andreessen Horowitz partner Olivia Moore says almost all consumer AI is really prosumer AI. We break down her TechCrunch interview and a16z’s seventh Top 100 Gen AI Consumer Apps report: the new spending data, why the top 1% of payers average $903 a month, how subscriptions crowd out ads and transaction fees, what ChatGPT’s ad revenue and Instinct’s volume imply, and which consumer categories AI has still not reached.

Read more
anthropic usage policy cruel behavior claude ai a boxing ring corner post with a towel over the ropes

Anthropic Bans ‘Cruel’ Behavior Against Its Claude AI

Anthropic has added a ban on “sustained and needless abusive or cruel behavior” toward Claude to its Usage Policy, effective 12 November 2026. We set out the exact wording, how Claude ending abusive chats enforces it, why Anthropic’s welfare research led here, what critics including Mustafa Suleyman and Pope Leo XIV say, every other change in the update, and what UK businesses using Claude should check.

Read more
anthropic interviewer 15 minute voice interviews a megaphone with a flared bell and a pistol grip

Anthropic invites users to 15-minute voice interviews via Anthropic Interviewer

Anthropic is prompting more Claude users to spend 15 minutes with Anthropic Interviewer, an AI that asks what they want from AI. The study runs until 6 October 2026, and for the first time participants can publish their full transcript. We explain what the invitation says, whether it is really a voice interview, who can take part, what happens to private answers, and why researchers have already re-identified people from earlier transcripts.

Read more
anthropic voice data claude ai model training a condenser microphone in a spider shock mount behind a pop filter

Anthropic prompts users to share voice data for AI model training

Claude users are seeing a new prompt asking them to let Anthropic train its AI models on their voice recordings. The setting is off by default and separate from the chat-training switch, but Anthropic’s own privacy pages still say it does not train on voice. We set out what the prompt says, how it compares with the 2025 chat change, OpenAI, Google, Meta and Amazon, the Illinois biometric lawsuit, and what businesses should do.

Read more
CHAT