Elon Musk Testifies in OpenAI Lawsuit

In his testimony for the lawsuit against OpenAI, Elon Musk made several claims, including that he didn't closely review the company's original term sheet and that OpenAI "stole a charity." He also clarified that his past use of the term "AI enabled robot army" was about safety, not a military force.

Chinese AI Lab DeepSeek Releases New V4 Models

Chinese AI lab DeepSeek has released preview versions of its new open-source models, DeepSeek-V4-Pro and DeepSeek-V4-Flash. The models feature large context windows and aim to compete with leading AI on performance benchmarks, particularly in coding and mathematics. The company also announced a 75% price cut for its V4-Pro model.

Anthropic's New 'Claude Design' Tool Creates Visuals from Text

Anthropic has launched Claude Design, an experimental product that allows users to create visuals like prototypes and slides from natural language descriptions. The tool, powered by the new Opus 4.7 model and developed in partnership with Canva, is available as a research preview for paying subscribers.

Anthropic Releases New Opus 4.7 AI Model and Security Tool

Anthropic has launched its new generally available AI model, Claude Opus 4.7, touting improvements in coding and image analysis. The company is also deploying a codebase-scanning security tool for businesses, which is powered by the new model.

Apple Reports Mac Shortages Driven by AI Demand

Apple is experiencing supply shortages for its Mac mini, Mac Studio, and Neo models, which the company attributes to a surprising surge in demand driven by AI-related needs. The base M4 Mac mini is sold out on Apple's website, leading to inflated prices on third-party sites like eBay.

Lawmakers Advance GUARD Act to Age-Gate AI Chatbots

The Senate Judiciary Committee has unanimously advanced the GUARD Act, a bill that would ban access to AI chatbots for individuals under 18 and require age verification for all users. The bill, co-sponsored by Senators Josh Hawley and Richard Blumenthal, will now proceed to the Senate floor.

Jared Birchall Testifies in Musk v. OpenAI Trial

Jared Birchall, who manages Elon Musk's family office, testified in the ongoing lawsuit between Musk and OpenAI. His testimony covered topics such as Musk's donations to the company, the contribution of four Teslas, and the decision to pause funding.

OpenAI Releases Limited-Access Cybersecurity Model GPT-5.5-Cyber

OpenAI is launching a specialized cybersecurity AI model called GPT-5.5-Cyber. Access to the tool will initially be restricted to a select group of trusted "critical cyber defenders" to ensure it is used securely and to prevent potential misuse.

Musk Testifies That xAI Used OpenAI Models to Train Grok

In court testimony, Elon Musk confirmed that his AI startup, xAI, used a technique called model distillation to train its Grok AI on OpenAI's models. Musk stated that such practices are common among all AI companies for validation purposes.

Google Rolls Out Gemini AI Assistant to Cars

Google is integrating its Gemini AI assistant into cars that are equipped with 'Google built-in,' replacing the existing Google Assistant. The update is designed to enable more natural, conversational interactions for drivers to perform tasks and get information hands-free.

OpenAI Instructs Codex Model to 'Never Talk About Goblins'

OpenAI has addressed a quirk in its Codex and GPT-5.1 models that caused them to frequently reference goblins and other fantasy creatures in unrelated conversations. The company added an explicit directive in the system prompt for the AI to avoid such references unless directly relevant to a user's query.

OpenAI Releases 'GPT-5.5' Model, Codenamed 'Spud'

OpenAI has released its most capable model, GPT-5.5, codenamed 'Spud,' which is available to paid subscribers. The model is said to offer enhanced autonomy for multi-step workflows and has shown improved performance in areas like coding and scientific research.