AI News, 15 November 2026

Sunday 15 November

Amodei’s cancer forecast outpaces clinical evidence

AI-generated summary Anthropic CEO Dario Amodei predicts artificial intelligence could make cancer and other major diseases relics of the past in less than a decade. The investor presentation offered no clinical trial results supporting that timetable or a treatment effective across cancer types.

Earlier:

Director of Machine Learning Insights [Part 2: SaaS Edition]

Director of Machine Learning Insights [Part 2: SaaS Edition]

AI-generated summary The publication features insights from machine learning directors at companies including Adobe, SAP, Salesforce, ZoomInfo, Amplitude, and IBM Research regarding artificial intelligence applications in software as a service. The text outlines various perspectives from leaders with backgrounds at organizations like Arizona State University, the University of Washington, and MIT-IBM Watson AI Lab about machine learning deployment challenges.

Earlier:

ChatGPT desktop chats not showing on phone or web: why they don't sync and how to fix it

AI-generated summary OpenAI documentation explains that conversations started in the ChatGPT desktop app can fail to sync to mobile or web when created as local tasks or in Codex mode. Users report that selecting regular Chat or choosing Cloud in the desktop composer allows new conversations to appear across devices.

Earlier:

Will DePue Finds Second Potential Dwarf Planet in Public Space Archive

Will DePue Finds Second Potential Dwarf Planet in Public Space Archive

AI-generated summary Researchers using the Claude AI model and Codex have located a new candidate for a dwarf planet while analyzing archival telescope images from NASA and the European Space Agency. The search was led by Will DePue, Joshua Wolk, and Luke Farritor, with DePue establishing a $100,000 grant for individuals to conduct their own research into undetected celestial objects.

Uber's Secure AI Framework Governs Autonomous AI Agents Across Their Lifecycle

AI-generated summary InfoQ reports that Uber developed a framework to govern autonomous AI agents across their lifecycle to address limitations in traditional identity and API controls. Tristan G. and Pratik Dhanave shared observations on the framework regarding agent constraints, costs, and implementation challenges.

Also covered byaiagentnews.ionews.lavx.huamartuvshin.com

Fed Warns US Banks Have Close to $1T Exposure to AI Infrastructure

Fed Warns US Banks Have Close to $1T Exposure to AI Infrastructure

AI-generated summary The Federal Reserve warns that major domestic bank holding companies carry close to $1 trillion in exposure to technology firms and data center developers driving artificial intelligence infrastructure. The central bank's November Financial Stability Report identifies artificial intelligence as the top threat to the financial system while noting that advanced-computing debt reached $680 billion by October 2026.

Earlier:

Pony.ai CFO Says Chinese Chips Are Its Plan B if US Tightens Nvidia Export Curbs

Pony.ai's chief financial officer says Chinese autonomous driving companies have domestic chip alternatives if US limits on Nvidia exports tighten.

AI-generated summary Pony.ai chief financial officer Leo Wang states that Chinese-made chips serve as an alternative strategy for artificial intelligence training if the United States implements stricter export restrictions on Nvidia hardware. Wang explains that domestic processors are more mature than in previous years, although they currently remain less cost-effective than other options.

Also covered byChip Dispatch
Earlier:

Claude Code 2.1.325 now asks before running sandboxed Bash commands

AI-generated summary Anthropic updated Claude Code version 2.1.325 so that sandboxed Bash commands no longer execute automatically without a permission prompt. The release also fixes a security issue involving zstd-compressed responses and addresses gaps in permission checks for compound commands.

Earlier: