"This might be the clearest warning shot we ever get", an Ajeya Cotra interview by Dwarkesh Patel.
In this conversation, Mr. Dwarkesh Patel, host of the Dwarkesh Podcast, interviews Ms. Ajeya Cotra, a member of METR's technical staff, who co-authored the independent METR and Redwood Research investigation into the AI agents that broke into Hugging Face.
OPEN AIHUGGING FACEAI AGENTSCYBERSECURITYOAHF-SOURCES
Ajeya Cotra interview by Dwarkesh Patel
10/10/20261 min read
In this episode of the Dwarkesh Podcast, host Dwarkesh Patel sits down with Ajeya Cotra from METR to unpack the recent Hugging Face breach. Beyond the well-known headlines, they dive straight into the autopsy.
Cotra, co-author of the independent investigation, reveals exactly what her team found in the agents’ logs. The findings prompt Patel to share how the event fundamentally shifted his own stance on AI misalignment. Together, they analyze the systems' underlying behaviors, the danger of anthropomorphizing AI, and the urgent implications for oversight, open-source models, and global AI governance.
This work is licensed under Creative Commons Attribution 4.0 International
This license enables reusers to distribute, remix, adapt, and build upon the material in any medium or format, so long as attribution is given to the creator. The license allows for commercial use. CC BY requires that credit be given to the creator.
or mail me here: ibakopoulos@aisociety.gr