AI News for April 10, 2026 — 7min.ai

[NEWS] Meta AI rockets from App Store rank 57 to 5 after Muse Spark launch; stock up 10%: The Meta AI app rocketed from No.
[NEWS] OpenAI launches $100/month ChatGPT Pro tier with 5x Codex usage limits: OpenAI introduced a new $100/month ChatGPT Pro plan offering five times the Codex usage of the $20 Plus tier, directly targeting Anthropic's identically priced Claude Max subscription.
[NEWS] Bessent and Powell summon Wall Street CEOs over Anthropic Mythos cyber risks: Treasury Secretary Scott Bessent and Fed Chair Jerome Powell assembled leaders of major US banks at Treasury headquarters to warn about cybersecurity risks posed by Anthropic's Claude Mythos model.
[NEWS] Amazon touts AI chip progress, hints at selling Trainium to third parties: Amazon CEO Andy Jassy used his annual shareholder letter to declare the company's custom AI chip business "on fire" and took a direct shot at Nvidia's dominance.
[RESEARCH] Anthropic's Mythos found a 27-year-old OpenBSD zero-day; company launches $100M Glasswing defensive coalition: VentureBeat's analysis of Anthropic's Mythos Preview red team results reveals the model autonomously found a 27-year-old bug in OpenBSD's TCP stack that had survived years of audits and fuzz testing — the specific run cost under $50.
[NEWS] CoreWeave strikes $21B AI computing deal with Meta through 2032: CoreWeave signed a $21B contract to supply AI computing power to Meta through 2032, deepening its relationship with the social media giant as Meta races to catch up in advanced AI model development.
[NEWS] OpenAI pauses Stargate UK data center project citing energy costs: OpenAI halted its Stargate AI infrastructure project in the UK as it reins in spending ahead of its anticipated IPO.
[TOOLS] Claude Managed Agents priced at $0.08/session hour; Notion, Rakuten, Sentry first to adopt: Anthropic’s Claude Managed Agents — the hosted infrastructure for autonomous AI agents announced yesterday — is priced at $0.08 per session hour on top of standard token costs.
[NEWS] Stealth Alibaba video AI model sweeps to top of global rankings on debut: A previously unknown video generation model developed by a team under Alibaba topped global benchmarks on its debut, sending ripples across China's AI industry.
[RESEARCH] Anthropic had a psychiatrist run 20 sessions with Mythos; system card says AI welfare 'increasingly concerns us': Anthropic’s 244-page system card for Claude Mythos — the company’s most capable but unreleased model — reveals the company commissioned a psychiatrist to conduct 20 hours of sessions with the model as part of its evaluation.

About 7min.ai · Subscribe to newsletter