[01]
Judge blocks Pentagon's Anthropic blacklist
A federal judge ruled the Pentagon illegally blacklisted Anthropic as a supply-chain risk, ordering the designation rescinded.
// daily-digest · fri 28 aug 2026 · morning refresh · last 72h focus
Ten items on Anthropic's legal and product news, AI lab talent dynamics, fintech infrastructure, and software engineering, plus a wildcard on supercentenarian immunology.
— refreshed for you, in 10 items.
A federal judge ruled the Pentagon illegally blacklisted Anthropic as a supply-chain risk, ordering the designation rescinded.
Anthropic previews an open specification letting AI agents safely operate lab and manufacturing hardware like microscopes and robotic arms.
Data shows Google DeepMind's share of elite AI research hires falling sharply as rivals including OpenAI and Anthropic gain ground.
OpenAI's first custom inference chip, built with Broadcom, beats Nvidia on performance per watt in independent benchmarks.
Reuters reporting detailed how Meta's leadership planned a 60% team-size cut across engineering, citing AI-driven efficiency gains.
RQD* Clearing raised a $74 million growth investment led by Bain Capital to expand clearing and custody infrastructure.
This week's Rails roundup covers a new HTML+ERB parsing engine, i18n scoping improvements, and Ractor-safety fixes.
Noah Smith imagines a 2029 scenario where a jailbroken language model helps design a bioweapon.
Sean Goedecke argues that adopting a professional persona at work doesn't necessarily require sacrificing one's authentic self.
Researchers found supercentenarians carry an unusually high proportion of rare, cancer-killing CD4 immune cells.