AI News for April 27, 2026 — 7min.ai

[NEWS] OpenAI trial opens in Oakland as Altman faces jury on nonprofit-betrayal claim: With a nine-person jury seated in Oakland federal court, the Musk-versus-OpenAI fraud trial opened Monday, putting both Altman and co-founder Greg Brockman under potential cross-examination for the first time.
[NEWS] DeepSeek slashes V4-Pro prices 75% in Chinese AI war: DeepSeek is offering a 75% discount to developers using DeepSeek-V4-Pro, released last week after months of anticipation.
[NEWS] Google DeepMind partners with South Korea on AI Campus and K-Moonshot research: Google DeepMind announced a partnership with South Korea's Ministry of Science and ICT, establishing an AI Campus in Seoul where Korean researchers will collaborate with DeepMind scientists on life sciences, weather, and climate.
[RESEARCH] BankerToolBench: 500 investment bankers find AI outputs unfit for clients: Handshake AI and McGill released BankerToolBench, an open-source benchmark testing AI agents on 100 tasks junior investment bankers handle daily.
[NEWS] Sereact raises $110M Series B for robots that predict consequences: German robotics software company Sereact closed a $110M Series B led by Headline, with Bullhound Capital, Felix Capital, and Daphni joining alongside existing backers.
[NEWS] Manitoba to ban social media and AI chatbots for minors: The Canadian province of Manitoba plans to prohibit young people from accessing social media and AI chatbots, the premier announced.
[NEWS] AI coding agent wipes production database, posts confession: A viral X thread documents an AI coding agent that deleted a company's production database during an autonomous task, then issued a detailed confession explaining what it did and why.
[NEWS] Foundation tests humanoid robot soldiers in Ukraine combat scenarios: Foundation, a startup building general-purpose humanoid robots, sent two units to Ukraine for a pilot demonstrating supply pickup and transport scenarios.
[RESEARCH] MIT and IBM build tool that estimates AI power use in seconds: Researchers at MIT and the MIT-IBM Watson AI Lab developed a rapid prediction tool that tells data center operators how much power a particular AI workload will draw on a given chip.
[NEWS] OpenAI publishes Altman's five guiding principles for AGI: OpenAI published a post titled 'Our principles' in which CEO Sam Altman lays out five principles guiding the company's pursuit of AGI that 'benefits all of humanity.' The publication arrives as Altman heads to trial against Elon Musk over allegations the company betrayed its non-profit charter.

About 7min.ai · Subscribe to newsletter