A lawsuit filed in the U.S. District Court for the Northern District of California accuses Anthropic, OpenAI, SpaceXAI and Google of illegally agreeing to slow their AI development in ways that harm consumers. Plaintiffs - four named subscribers to paid services including ChatGPT, Claude, Grok and Gemini, seeking to represent a nationwide class of paid users - contend the companies coordinated on September 12 when Anthropic CEO Dario Amodei published an essay calling for industrywide deceleration for safety and leaders including Sam Altman, Elon Musk and Demis Hassabis publicly echoed support. Lawyers argue that an agreement among top rivals to advance more slowly than competitive pressures would dictate has an anticompetitive effect and reduces the value of paid subscriptions.
The complaint highlights Amodei’s own recognition that such safety talks could raise antitrust issues and his proposal that the government mediate or grant a narrow waiver to allow cross-lab safety discussions. OpenAI’s response welcomed a federal safety framework while saying it would begin work without waiting for legislative exemptions. Plaintiffs’ counsel warns that private, self-interested pacts on AI safety risk concentrating control and endangering consumers; company representatives had not commented at the time of filing.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
OpenAI and Anthropic exaggerated the severity of AI security breaches to influence government protection of their markets. Insiders suggest the claims were overstated to pressure federal regulators and secure industry turf. (nypost.com)
ZK-JPEG is a cryptographic tool that proves an image was correctly compressed from a secret input using zero-knowledge proofs, even after lossy JPEG encoding. It also supports verifying a wide range of image transformations with minimal cost, enhancing image authenticity and integrity. (eprint.iacr.org)
A Microsoft director described AI data scraping as the largest theft of labor in human history. The controversy involves AI companies collecting vast amounts of data without proper compensation or consent. (tomshardware.com)
AI coding assistants have led many developers to rely on agents rather than writing code themselves, raising concerns about their actual coding skills. Interviewers are debating whether traditional testing methods remain effective or need adaptation in this new AI-driven landscape. (news.ycombinator.com)
Irregular reports that a single underlying issue caused four AI lab breaches. The breaches were linked to a common vulnerability in the labs' security systems. (thenextweb.com)
Anthropic, OpenAI, SpaceXAI, and Google are being sued over their calls to accelerate AI development. The lawsuit questions the safety and ethical considerations of rushing AI progress. (politico.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.