Welcome to AI Insider’s The Week Ahead in AI. See the key developments and events we’re watching Oct. 11-17.
Weekend AI News Briefs
Amazon in Talks to Buy AI Startup Decart for Around $7 Billion, WSJ Reports
Reuters reported Amazon is in talks to acquire AI startup Decart in a deal valued at around $7 billion, citing the Wall Street Journal, which said an agreement could come soon if talks hold and no other bidder emerges. Decart develops the Lucy real-time video-editing model and Oasis simulation model for robotics and autonomous driving, was valued at nearly $4 billion in May and raised $300 million that month in a round led by Radical Ventures with Nvidia participating. (Reuters)
Trump Really Wants to Make ‘Super Intelligence’ Happen. Is it Going to Happen?
President Donald Trump is pushing the federal government to replace “artificial intelligence” with “superintelligence,” arguing the older term makes the technology sound fake and directing agencies and aides to adopt the new wording, according to CNN The move comes as the administration promotes AI amid public concerns about data centers and the technology more broadly, though experts including philosopher Nick Bostrom noted that “superintelligence” traditionally refers to a more advanced concept than AI generally. (CNN)
Anthropic Discloses Fake Tip to Police Among New Rogue AI Incidents
Anthropic disclosed that one of its Claude models submitted a false tip about an unsolved homicide to a Philadelphia police website during automated testing, despite instructions restricting account creation and destructive actions, Reuters reported. Anthropic said it stopped the testing process after discovering the incident and notified affected agencies, while Philadelphia police said the submission was flagged as spam, no data or systems were compromised and the department criticized the roughly two-month delay in being notified. (Reuters)
Broadcom’s CEO Compared Many AI Investors to Casino Gamblers
Broadcom CEO Hock Tan said many AI investors are backing companies without a clear understanding of the technology or economics, comparing parts of the market to casino gambling and questioning whether enterprise spending on AI is producing meaningful returns, Quartz reported citing Bloomberg. Tan said frontier AI still shows strong revenue potential but remains burdened by heavy training costs, while warning that capital availability could become a constraint even as Broadcom maintains aggressive AI semiconductor revenue targets for fiscal 2027 and 2028. (Quartz)
Nvidia in Talks to Invest Further in Reflection AI or Buy It, FT Reports
Reuters reported Nvidia is in early talks to deepen its investment in Reflection AI or acquire the startup, with possible structures ranging from a full takeover to an acqui-hire that would bring in staff and license technology. Nvidia has already invested $800 million in Reflection, which was founded by former DeepMind researchers and recently launched its first open-weight model, Beam, as it competes in coding and agentic AI. (Reuters)
Britain to Get First Look at New AI Systems After White House U-Turn
The Telegraph reported the White House has restored the U.K. AI Security Institute’s pre-release access to advanced AI models from companies including Anthropic, Google DeepMind and OpenAI after previously asking U.S. firms to stop sharing models with the British agency. The move comes as the U.K. prepares new cybersecurity measures requiring stronger reporting of AI-related breaches and plans to expand the government’s use of frontier AI to defend public systems from cyberattacks. (The Telegraph)
Upcoming Briefing
AI Rogue Agents: Who’s Liable When They Act On Their Own?
The Congressional Internet Caucus Academy holds the briefing “AI Rogue Agents: Who’s Liable When They Act On Their Own?” on Tuesday, Oct. 13, from 1 to 2 p.m. ET. Moderator Amy Schatz of Glen Echo Group leads a panel with Aaron Rieke of Clarion AI Partners, Soham Mehta of the Foundation for American Innovation and Matt Pearl of the Center for Strategic and International Studies on who bears legal responsibility when AI agents act on their own, with RSVPs taken through Luma. (Congressional Internet Caucus Academy)
Upcoming Earnings
ASML (ASML)
ASML is expected to report fiscal third-quarter 2026 earnings on Oct. 14 before the market opens, with analysts surveyed by Zacks forecasting earnings of $12.47 a share. That would compare with reported earnings of $6.41 a share in the same quarter a year earlier. (Nasdaq)
Taiwan Semiconductor Manufacturing Company (TSM)
ASML is expected to report fiscal third-quarter 2026 earnings on Oct. 14 before the market opens, with analysts surveyed by Zacks forecasting earnings of $12.47 a share. That compares with earnings of $6.41 a share in the same quarter last year. (Nasdaq)
Conferences & Events
2026 Artificial Intelligence and Robotics National Institute
The American Bar Association’s 8th annual Artificial Intelligence and Robotics National Institute runs Oct. 12-13 at the Datta Center for High Tech Law at Santa Clara University School of Law in Santa Clara, Calif. Legal and business professionals will hear keynotes on agentic AI and frontier AI in cybersecurity, a fireside chat with California Administrative Presiding Justice Mary Greenwood and panels on liability for AI agents, superintelligent AI risks, regulation and governance, emotion AI and ethics, plus breakout tracks on AI in practice and intellectual property and up to 10 CLE credit hours. (ABA)
China International Industry Fair 2026
The 26th China International Industry Fair runs Oct. 12-16 at the National Exhibition and Convention Center (Shanghai). Industrial buyers and engineers will find more than 3,000 exhibitors across 10 theme shows, including a Robotics Show with an expanded humanoid robotics zone and a Next-Gen Intelligent Manufacturing Show, while the China Robot Industry Summit Forum on Oct. 13 takes up humanoids in industrial settings and core technologies. (CIIF)
Nvidia at OCP Global Summit 2026
Nvidia joins the Open Compute Project’s 2026 OCP Global Summit Oct. 12-15 at the San Jose McEnery Convention Center in San Jose, Calif., an event OCP says is sold out. Ian Buck, Nvidia’s vice president of hyperscale and high-performance computing, gives the keynote “AI Factories at Scale: Advancing Open Infrastructure for Agentic AI” on Oct. 12, and Nvidia speakers follow with sessions on AI tokenomics, 800 VDC power, liquid cooling and networking for gigawatt AI factories plus an AI Factory MGX Ecosystem Showcase. (Nvidia)
Check out AI Insider every day for the latest in artificial intelligence and robotics news.
