DailyAI Briefing: Fed survey names AI top financial risk, plus 4 more
Banks and investors now name AI as the biggest danger to markets in a Fed survey, while borrowing for data centres soars. Also today: officials say an AI tool did not cause the halt to US flights, OpenAI splits up its robotics team, Amazon plans to collect data on every product sold online, and Nvidia staff push for the return of Jensen Huang.

Welcome to the DailyAI Briefing, the day’s five biggest AI stories.
Let’s dig in.
AI tops the list of financial risks in a Fed survey
87%. That is the share of contacts in a New York Fed survey who named artificial intelligence as a risk to the US financial system, up from 50% in the spring. The figures appear in the central bank’s Financial Stability Report, released on 13 November. The 23 people who answered work at banks, broker-dealers, investment funds and advisory firms, and the report stresses that their answers are not the view of the Fed itself.
They had three worries: share prices that could fall if AI earnings disappoint, cyberattacks by state hackers or by frontier models themselves, and complicated deals used to pay for the buildout. The next most cited risk, a broad market correction, was named by 52%.
Now available: The Financial Stability report which presents the Federal Reserve Board’s current assessment of the stability of the U.S. financial system. Financial stability supports the objectives assigned to the Federal Reserve, including full employment and stable prices, https://t.co/avC0vmutgr
— Federal Reserve (@federalreserve) November 13, 2026
The report’s own numbers explain the nerves. Debt raised for advanced computing passed $680 billion by October, about five times as much as in all of 2023, and it now makes up about 31% of the investment-grade bonds sold by non-financial companies this year. A survey of 23 people is not a forecast. Still, when the people who lend to the industry put it at the top of their list of worries, the money for the next data centre could get more expensive.
Officials say AI did not cause the nationwide halt to US flights
12,462. That is roughly how many flights were cancelled on Friday after a fault in the air traffic control network forced a nationwide ground stop. The aviation regulator published its account of the outage on 15 November. It says old equipment became unstable when data flooded the network at 1,000 connection attempts every four to five seconds. The flood was traced to a radar site in Texas and a radio site in Arkansas, where the network provider had been installing a new router. Departures resumed after more than four hours.
There has been a temporary telecommunications outage – out of an abundance of caution all flights have been halted. Every flight has been accounted for and international flights will be handled on a case by case basis.
— The FAA (@FAANews) November 13, 2026
Questions had quickly turned to SMART, a recently introduced AI planning tool. Transportation Secretary Sean Duffy said the tool only makes suggestions to controllers and is not connected to the radar and flight data systems, and the regulator now says the two “are not related”. When something big breaks, the newest AI system is the first suspect. Teams that keep a person in charge of decisions, and log what the model suggested, can answer that question in minutes.
OpenAI breaks up its stand-alone robotics team
OpenAI has moved its robotics researchers into its main research group under chief research officer Mark Chen, and its applied robotics and hardware staff into the Applied organisation run by president Greg Brockman. The reorganisation was first reported on 13 November, and the company confirmed it. OpenAI says robotics remains a core part of its long-term plans and that the change is meant to speed the work up.
Want AI to save you hours a week? Our free playbook shows you how.Get the free guide →The reasoning is the interesting part. At the start of the year, OpenAI expected progress to come from building on video models such as Sora. Since then, it says, general models such as GPT-6 Astra have shown they can be applied directly to robots. The change came so late that the company’s booth at CoRL, a leading robotics conference in Austin this week, sat empty.
If a general model can learn to control a robot arm, separate robotics teams start to look unnecessary. That is a bet, not a result. OpenAI has yet to show its own humanoid robot, and it is aiming for a first public demonstration in 2027.
Amazon plans to collect data on every product sold online
1 October 2027. That is the release date written on an internal Amazon document describing World Product Graph, a plan to collect data on almost every item for sale anywhere, whether Amazon stocks it or not. Details of the plan were reported on 13 November by a news agency that reviewed the document. Amazon has already signed up some brands, including Coca-Cola and Bayer, to send it data, and the plan also has Amazon crawl the websites of other retailers.
The document admits that many answers from Rufus, the shopping assistant on Amazon, “are factually wrong or incomplete”. Better data would fix that, and it would also give AI shopping agents a reason to look on Amazon first. Amazon said brainstorming documents like this one do not necessarily reflect its current plans.
Our report on AI assistants that can now pay for purchases shows where this is heading. Once software does the shopping, whoever holds the best product data decides what ends up on the shelf.
More than 115,000 people ask Nvidia to bring back Jensen Huang
5,965 current Nvidia employees had signed an open letter by Saturday afternoon, asking the board to reinstate Jensen Huang. More than 115,000 people have signed it in total. Several well-known technology bosses have also signed the letter, according to a report on 14 November. They include Elon Musk, Tim Sweeney and Garry Tan. The letter says Huang was removed “without meaningful due process” and asks the board to bring him back while an independent investigation establishes the facts.
Huang resigned on Thursday, days after admitting an affair with a chip executive whom US intelligence reportedly suspects of being a Chinese spy. Finance chief Colette Kress is now interim chief executive. Our earlier report covers how the shares reacted to his exit.
The letter is not endorsed by Nvidia or by Huang, and petitions rarely change a board’s mind. But open dissent among staff at the company that makes most of the chips AI runs on is something every one of its customers will be watching.
That’s all for today. See you tomorrow.


