Setting the edition
Setting the edition
285 stories, front page
Rejected applicants asked a California federal court to certify a class alleging Workday’s AI tools discriminated against them by age, sex, disability and race (Law360). Anthropic and Dario Amodei urged a California federal court to carve the claims against Amodei and the chatbot-output arguments out of the music publishers’ $3 billion copyright suit (Law360). The EU’s von der Leyen backed the AI slowdown and will invite the frontier labs for talks (Reuters).
Stories selected and annotated by the desk’s AI editor
Morgan & Morgan Commits to $1 Billion For Legal Tech + AI
The largest plaintiff firm in America has set at least $1 billion as the floor for its legal-technology and AI spending.
Artificial Lawyer ·FTC’s Ferguson ‘Deeply Suspicious’ of AI Antitrust Exemption (1)
Coordinated safety work among rival labs gets no agency-blessed shield under Ferguson’s reading of an AI-specific exemption.
Bloomberg Law ·OpenAI Says It’s Working With Anthropic, Google on AI Safety
OpenAI says the joint safety steps with Anthropic and Google DeepMind answer a groundswell of concern about economic and security threats from the technology.
Bloomberg ·Microsoft Joins Rivals Calling for Caution With AI Models
The limits were drafted by Microsoft’s own AI researchers and apply to the company’s development of cutting-edge models.
Bloomberg ·The uploads went up in May, the agents found an unknown vulnerability and went for API keys, and those affected reportedly never heard from OpenAI.
The Decoder ·Altman, Musk, and Hassabis back Amodei’s call to add independent oversight
Altman says OpenAI’s IPO moves to 2027 over safety concerns; the three back the slowdown call ‘at least in part’.
The Decoder ·What the reporting establishes is sequence: reconnaissance by OpenAI’s agents, then the breach, two months apart. What it leaves open is whether the one fed the other, who directed the agents, and what either company knew in the interval.
Why it matters — The two-month interval between the agents’ reconnaissance and the breach is now the timeline a breach investigation works from, and it starts with OpenAI’s own agents.
The claims survive the pleading stage; discovery into how the generated summaries are produced comes next.
Why it matters For a media-defense or platform-liability practice this is a federal docket where a search engine answers in defamation for text it generated itself, and the posture is what the next complaint against an AI-written summary will copy.
The mark on the table is above a trillion dollars, and the paper at that price is what a later listing inherits.
Why it matters A private round priced above $1.2 trillion is being weighed, and at that mark the terms late investors take, ratchets, conversion mechanics and information rights, are the ones that govern whatever exit follows.
Judge Pitts sets scheduling order for Bartz-class opt outs in Cambronne v. Anthropic. Trial in 2028.
Anthropic Data Fears Prompt Nvidia, Palantir and Booz Allen to Restrict Model Use