The BasePodcastDecoder with Nilay Patel71 min

The OpenAI board member who fired Sam Altman says we still can't trust AI CEOs

In a 71-minute Decoder interview, Nilay Patel talks with Helen Toner of Georgetown's CSET, a former OpenAI board member. Toner argues that getting AI to do what users want is not on track to be solved, proposes a higher safety bar before training new models, and says shared fear of losing control could bring the US and China closer.

Open the source ↗5 quotations checked, word for word ↓
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

Helen Toner, a former OpenAI board member, says the future should not hinge on trusting any single AI executive, and calls for more people and structures involved.

For lawmakers

Helen Toner of CSET suggests lawmakers set a required level of confidence in safety before AI companies train or release new models, instead of a fixed pause.

For researchers

Helen Toner says lab staff describe a large backlog of interpretability, evaluation-awareness and control experiments they lack time to run between model releases.

The question it raises

Helen Toner asks whether a technology its builders may not control could give rival powers like the US and China a shared reason to cooperate.

In their own words

Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.

From our own transcript of the episode’s audio, made by machine. Speaker names come from what is said, not from a voice match.

“that's why I do the work that I do because I think we need to have different structures, different, many more people involved, many more perspectives being represented”
Helen Toner, Executive Director, Center for Security and Emerging Technology, Georgetown UniversityAsked by Nilay Patel whether AI CEOs, including Sam Altman, are trustworthy, after saying hopes for the future should not depend on one individual.
Found word for word in the sourceSpeaker known from context, not confirmed by listeningspoken clean-up: that's (said twice)
“I think the most coherent one is trying to get the AI to do what the user wanted it to do, and it seems like that is as far as I can tell, not on track to be solved”
Helen Toner, Executive Director, Center for Security and Emerging Technology, Georgetown UniversityToner on the different meanings of 'alignment', naming the one Toner finds most coherent.
Found word for word in the sourceSpeaker known from context, not confirmed by listening
“when the people who are building the systems tell us, oh, we have to solve this within twelve to twenty four months. I think that's terrifying, and I hope they're wrong.”
Helen Toner, Executive Director, Center for Security and Emerging Technology, Georgetown UniversityToner on how much time there is to solve the science of controlling advanced AI; Toner adds there is probably longer.
Found word for word in the sourceSpeaker known from context, not confirmed by listeningstarts inside a longer claim, right after “and”
“Maybe the best way to do that is just to like set a higher bar for here's how confident we need to be in the safety of these systems before we proceed with something, whether it's releasing a new model or I think more importantly, like even training a new model”
Helen Toner, Executive Director, Center for Security and Emerging Technology, Georgetown UniversityToner, after saying a fixed six-month pause would likely go badly, on how to give safety work more time.
Found word for word in the sourceSpeaker known from context, not confirmed by listening
“if that is actually what we think could seriously happen then China doesn't want that either. And so then it becomes a question of okay, how do we try and get them on the same page of this threat that we're worried about”
Helen Toner, Executive Director, Center for Security and Emerging Technology, Georgetown UniversityToner on the 'but China' argument, after saying many people are not yet convinced AI could take over from humans.
Found word for word in the sourceSpeaker known from context, not confirmed by listeningstarts inside a longer claim, right after “but”

The record

Questions
Are AI companies putting stronger safeguards into practice? · Who can demand evidence, delay a release or stop an unsafe deployment? · Can AI systems act beyond what people intended or allowed?
People
Helen Toner (Speaks, Executive Director, Center for Security and Emerging Technology, Georgetown University) · Nilay Patel (Speaks) · Sam Altman (About them)
Also covered by
The Verge
What we read
We read the whole source.
Found by
The Radar, run of November 15, 2026, via pod-decoder-with-nilay-patel (tier 1: the source’s own channel or an official record)
Cite this
Decoder with Nilay Patel, “The OpenAI board member who fired Sam Altman says we still can't trust AI CEOs”, November 14, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/decoder-2026-11-14-helen-toner-ai-ceos-trust
Our notesOnly one STT transcript; the second transcript failed, so quotes are not cross-checked against a second one. The transcript has no speaker labels (all '?'); attribution to Toner is from the flow of the interview, hence speakerCertain false. The episode's headline line ('if our hopes for the future depend on any one person being trustworthy') is garbled in the STT ('if we have to hopes...'), so it is not quoted; The Verge publishes an official transcript at the also URL that a person could compare. Patel says the talk was recorded the day after the midterm elections and before Nvidia opened its investigation of Jensen Huang; no exact recording date given. Transcript includes ads and a CrowdStrike sponsored segment, not used. Discussion touches Anthropic's S-1 and constitution, but the entry is not about Anthropic.