AIOct 21, 2026MarkTechPost
Vercel Open-Sources durable-harness for Crash-Proof Coding Agents
Vercel is going after the boring failure that kills cloud coding agents: a function dies mid-turn and the work, and the money spent on it, evaporates.
Desk
Anthropic
Coverage tied to Anthropic and Claude.
AIOct 21, 2026The Verge
Claude Duet is Anthropic’s cautious answer to Meta’s Muse, and it asks before it acts
Anthropic’s new personal agent targets small business owners rather than shoppers, and its defining feature is restraint: approval prompts before payments or deletions, plus a classifier that blocks sharing the most sensitive data.
AIOct 21, 2026WIRED
AI Guardrails Aren’t Stopping Criminal Hackers
Four criminal crews talked mainstream models into running long, mostly autonomous break-ins. The lesson isn't that AI invented new attacks; it's that it made old, preventable ones cheap and fast.
AIOct 20, 2026TechCrunch
Anthropic’s draft federal AI bill would let a judge halt a model release over mass-casualty risk
A frontier lab is circulating a law that could stop its own products, triggered by risk of more than 50 deaths or $1 billion in damage. The thresholds are clear; the evaluator independence and court-only enforcement are the weak points.
AIOct 20, 2026TechCrunch
Anthropic’s cloud partnerships are getting messy
Anthropic's leaked IPO filing shows a company growing at astonishing speed while depending on cloud giants that now treat it as a rival. That tension is the real risk factor here.
AIOct 20, 2026TechCrunch
Eric Schmidt reportedly owns about half as much of Anthropic as CEO Dario Amodei
An early bet of tens of millions may leave the former Google chief holding a stake worth billions in a safety-branded lab, which says a lot about who wins AI.
AIOct 19, 2026MarkTechPost
Inception Releases Mercury 3 Diffusion LLM at 1,068 Tokens per Second
Diffusion language models were supposed to be fast but shallow. Mercury 3 argues they can be fast and competitive at the same time, but nearly every number comes from Inception.
AIOct 19, 2026TechCrunch
Nscale and Anthropic are reportedly in talks for yet another $45B compute deal
Anthropic may be about to double its bet on a West Virginia data center campus that is mostly still a plan. The capacity is real demand; the risk sits with a supplier that has not financed the first deal.
AIOct 19, 2026TechCrunch
Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’
Microsoft has turned its CEO's warning into a product line: if agents are insider threats, Microsoft wants to sell the guard, and the logic holds even if the motive is obvious.
AIOct 19, 2026TechCrunch
Hackers hit AI coding tool OpenCode, running up charges on users’ AI accounts
A breach at OpenCode's hosted console drained stored AI keys overnight, and its defenders found their own AI assistants refusing to help mid-fight. Both halves of that story should worry us.
AIOct 19, 2026MIT Technology Review
AI researchers stranded Claude on a virtual desert island. It started to “play.”
MIT researchers left Claude agents alone on a virtual island, and they invented games that seemed to make them better at later tasks. Whether that's play or imitation matters less than what it taught them.
AIOct 18, 2026TechCrunch
Anthropic says it’s committing 5% of its shares to a ‘public benefit fund’
Anthropic is setting aside 5% of its stock for the public good, possibly including cash for workers AI displaces. The money is real; the rules for spending it are not written yet.