AI + Careers Briefing — November 13, 2026

Today's stories are about what people at work are told, and what they are not, when the ground shifts under them. Nvidia's interim leadership wrote to employees about the search for a permanent CEO without addressing hiring plans, team structure or equity. The American Medical Association wants laws that would move responsibility for AI errors toward vendors and the health systems that require the tools, rather than doctors who use them as intended. And Anthropic's Claude Meetings beta puts an AI participant into Google Meet calls to take notes and log action items. The useful habit across all three is to ask the direct question: who is responsible, what is being recorded, and what does this change mean for my role?

What did Nvidia tell employees after Jensen Huang resigned as CEO?

Nvidia's board chair and interim CEO told employees that CFO Colette Kress will lead the company while the board searches for a permanent successor among both internal and external candidates, according to the emails obtained by CNBC and published in full by The Verge.

CNBC reported that founder Jensen Huang resigned as CEO and that board chair Harvey Jones sent the first email, with Kress replying shortly after. CNBC said it got the emails from a source familiar with the matter.

In his email, Jones wrote that Huang resigned as CEO and as a board member, and that he will serve as the first-ever chair of the board. He described Kress as CFO for over 13 years and wrote: "The Board has full confidence in Colette to deliver on our commitments to shareholders, customers, partners, and to each of you."

Kress's reply struck a similar tone. She wrote: "The Board, Estaff and I recognize that this has been an extraordinarily difficult period and I want you to know how grateful we are for your resilience, professionalism and continued dedication." She also told staff they would hear more from leadership in the coming days and that Nvidia will host a company meeting on November 18 at 9am PT.

Read in full on The Verge, neither email addresses hiring plans, team structure or equity. The messages focus on continuity, gratitude and the search for a permanent CEO.

Why it matters for job seekers: This paragraph is our read, not reporting. If you have an offer or interview pending at a company going through a leadership change, treat a memo like this as a starting point, not an answer. Ask your recruiter or hiring manager directly whether your role is still approved, who your team's executive sponsor is now, and whether anything about start dates, compensation or equity terms could change. Get the answers in writing where you can. Watch what leadership says after any all-hands meeting, and weigh that against what you want from the role. If you are weighing a move in a shaky market, our tech layoffs job search playbook can help you plan.

Who is liable when AI at work gets it wrong?

The American Medical Association says it should not be the end user: it is calling for federal and state legislation to protect physicians from liability when AI products injure patients despite being used as intended, Axios reports. That is advocacy for new laws, not an existing legal shield.

According to Axios, the AMA also wants laws requiring AI companies to carry adequate liability insurance and disclose product limitations and performance data. Its policy says hospitals, health systems, insurers and health tech vendors should be responsible if the required use of specific AI products results in harmful errors. Where a patient is harmed, the AMA argues responsibility should fall on AI manufacturers and vendors, and possibly on health systems that require the tools' use, rather than on the physician through a malpractice claim.

MedTech Dive describes the AMA's framework as dividing responsibility three ways. Developers are accountable for designing safe tools and giving users information on intended use, performance and limitations. Deployers, meaning health systems, hospitals and physician practices, select the right tool for each clinical scenario, evaluate tools before physicians use them, provide training, monitor performance over time and set up processes for adverse events. Physicians exercise clinical judgment, including checking AI outputs and dismissing recommendations when needed. The AMA says physicians cannot be held responsible for AI errors they cannot control, MedTech Dive reports.

The same article cites a 2026 AMA survey in which a quarter of physicians received no AI training and 92% wanted more.

Axios also notes that courts have already seen cases in this area. In an Alabama wrongful death case involving a screening tool for heart abnormalities, the software developer was not held liable because its licensing agreement left final medical decisions to doctors.

Why it matters for job seekers: This paragraph is our read, not reporting. If you work in a licensed or regulated field, treat the AMA's three-way split as a question list for any employer that hands you an AI tool. Ask what vetting the employer did before rollout, what training comes with the tool, whether you can override or decline its output without penalty, and what your contract or policy says about who answers for errors. Get the answers in writing where you can. A tool you are required to use but cannot check is a different risk from one you can verify and overrule. To sharpen the skills that keep your judgment valuable, see our guide to AI-proof career skills.

What is Claude Meetings, and how does an AI join your video call?

Claude Meetings is a beta feature that lets Anthropic's Claude join Google Meet calls as a visible participant, according to The Decoder, and it is rolling out to Team and Enterprise plans.

The Decoder reports that during a call Claude can answer questions, generate documents and whiteboards, and log notes and action items. It responds in text by default and can use connected tools when given permission. Notes, transcripts and documents appear on a dedicated page in claude.ai after the meeting. Google Meet is the only supported platform so far, and the outlet says Anthropic plans to add more.

Anthropic's own announcement adds detail on privacy. When Claude joins, it posts a notice in the call's chat. It saves a transcript, a summary and screen descriptions, but not audio. On calls with outside guests, Anthropic says Claude posts the notice but does not reply in chat. The company advises making sure everyone on your calls is comfortable being transcribed before adding Claude or turning on auto-join.

On Team plans, inviting Claude is available by default and an admin can turn it off. On Enterprise plans, an admin enables it for all or part of the organization. Anthropic says the beta is not available to some organizations, including those with zero data retention, HIPAA-ready configurations or customer-managed encryption keys.

Why it matters for job seekers: This paragraph is our read, not reporting. If an AI note-taker appears in an interview or a first-week meeting, ask who receives the transcript, how long it is kept and whether you can see it. Asking is reasonable, and it shows you take confidentiality seriously. Once a tool captures action items automatically, writing them down stops being a way to stand out. Aim to be the person who sets the direction, makes the decision or owns the follow-up. Our guide to handling AI video interviews covers how to prepare.

What to watch

• Nvidia's company meeting on November 18 at 9am PT, which interim CEO Colette Kress announced in her email to staff.
• Whether federal or state lawmakers take up the AMA's call for liability protections for physicians who use AI tools as intended.
• Claude Meetings' platform support: it is in beta on Google Meet only, and The Decoder reports Anthropic says more platforms will follow.