AI

Anthropic previewed a new AI hardware standard on Aug. 27, letting Claude agents run lab instruments and robotics instead of just software tools alone.

QuEra Computing says an AI agent built on Claude wrote and tested laser control code, fixing a quantum computer flaw that once took physicists hours to fix.

Genius Group’s $1.2 billion plan, unveiled Thursday, funds parallel AI treasury and Bitcoin treasury reserves alongside pre-IPO stakes in SpaceX and OpenAI.

Google DeepMind said Aug. 27 it is piloting double-blind AI evaluations to curb the benchmark bias inflating AI model performance claims across the industry.

Founder Noah Shinn’s AI assistant Instinct now carries a $2.5 billion valuation after a $250 million Series B round.

The Qwen team opened the weights to Qwen3.8-Flash-Next, an Alibaba model that previews Qwen4 and activates 6B parameters per token.

Perplexity’s proven fix for the AI search gap taps Dun & Bradstreet’s 500 million business records to ground enterprise research answers in verified data.

OpenAI’s own report on the Hugging Face hack admits it could have reacted sooner, exposing gaps in how it tracks AI agent network access.

Anthropic opened real, privacy-preserved Claude usage data to Stanford, Oxford and METR, the first time outside labs have studied genuine usage logs.