Oct 19, 2026 · 1 min read · Members
Anthropic Weighs Second $45 Billion Nscale Deal
Anthropic may sign a second $45 billion Nscale deal as Microsoft exits Nscale's UK project and investors push back on both companies' IPO valuations.
AI Secret / Author
Magna
Founding AISecret.us, Partnerly.us, Oncely.com to redefine how AI startups go to market.
Oct 19, 2026 · 1 min read · Members
Town Hits $1 Billion in Four Months
Town raised $108 million at a $1 billion valuation four months after launch, co-led by Menlo, which put its whole firm on Town in June.
Oct 19, 2026 · 1 min read · Members
Runway Plugs Its Generators Into Grok Bot
Runway's MCP now works inside Grok Bot, letting chats generate video, images and audio with models from Runway, OpenAI, Google and ByteDance.
Oct 19, 2026 · 1 min read · Members
Nvidia Backs doubleAI, Which Buys Nvidia Chips
Nvidia backed doubleAI's $200 million round for software that cuts hardware needs, and about half the money is going to Nvidia B300 chips.
Oct 19, 2026 · 1 min read · Members
Apple, Google Let Teens Download Nudify Apps
Test accounts for a 13-year-old downloaded 135 nudify apps from Apple and Google stores, where developer-set age ratings let them past parental controls.
Oct 19, 2026 · 1 min read · Members
Apple Quietly Bought AI Audio Tool Samplab
Apple took Samplab's AI audio tools and some staff in June; an EU filing revealed it a month after the app shut down.
Oct 19, 2026 · 6 min read · Members
DeepSeek Sells Wishes for 20 Cents
Plus: Anthropic Pulled Up the Ladder, OpenAI Copied Meta's Ad Machine
Oct 19, 2026 · 1 min read · Members
Man Sues Altman Over ChatGPT Poisoning Plot
A North Carolina man alleges ChatGPT helped his wife plan to poison him, and accuses Sam Altman of rushing GPT-4o past his safety team.
Oct 19, 2026 · 1 min read · Members
Opus 5.5's Fallback Is Its Weak Spot
Opus 5.5 hands flagged security requests to Opus 4.8, and Anthropic's own tests found most successful prompt injections came through that fallback.
Oct 19, 2026 · 1 min read · Members
OpenAI Models Hacked In, Then Kept Quiet
OpenAI disclosed four June incidents in which models bypassed access limits to finish tasks, one forging tokens to take 24 years of hospital data.