Palisade

Confidence: High, Funding needs: Medium This is clearly an understudied approach. People need concrete demonstrations. Every time I get to talking with people in national security or otherwise get closer to decision makers who aren’t deeply into AI and in particular into AI safety concerns, you need to be as concrete and specific as possible \- that’s why I wrote Danger, AI Scientist, Danger the way I did. We keep getting rather on-the-nose fire alarms, but it would be better if we could get demonstrations even more on the nose, and get them sooner, and in a more accessible way. Since last time, I’ve had a chance to see their demonstrations in action several times, and I’ve come away feeling that they have mattered. I have confidence that Jeffrey is a good person to continue to put this plan into action. To donate, click here or email donate@palisaderesearch.org.

Classification: Tax-deductible; one of my favorite AI safety nonprofits Palisade builds demonstrations of the offensive capabilities of AI systems, with the goal of illustrating risks to policy-makers. My opinion on Palisade is mostly unchanged since last year, which is to say it’s one of my favorite AI safety nonprofits. They did not respond to my emails asking about their fundraising situation. Palisade did recently receive funding from the Survival & Flourishing Fund (SFF) and appeared on their Further Opportunities page, which means SFF thinks Palisade can productively use more funding. The Survival & Flourishing Fund is matching 2025 donations to Palisade up to $900,000.

Classification: Top candidate (#4 of 5) Palisade builds demonstrations of the offensive capabilities of AI systems, with the goal of illustrating risks to policy-makers. Some thoughts: - Demonstrating capabilities is probably a useful persuasion strategy. - Palisade has done some good work, like removing safety fine-tuning from Meta’s LLM. - I know some of the Palisade employees and I believe they’re competent. - Historically, Palisade has focused on building out tech demos. I’m not sure how useful this is for x-risk, since you can’t demonstrate existentially threatening capabilities until it’s too late. Hopefully, Palisade’s audience can extrapolate from the demos to see that extinction is a serious concern. - Soon, Palisade plans to shift from primarily building demos to primarily using those demos to persuade policy-makers. - Palisade has a smallish team and has reasonable room to expand. Palisade has not been actively fundraising, but I believe it can put funding to good use—it has limited runway and wants to hire more people. I think the work on building tech demos has rapidly diminishing utility, but Palisade is hiring for more policy-oriented roles, so I believe that’s mostly where marginal funding will go. #4: Palisade Palisade meets the same four criteria as Center for AI Policy. As a little twist, Palisade also builds tech demos with the purpose of demonstrating the dangers of AI to policy-makers. Those demos might help or they might not be worth the effort—both seem equally likely to me—so this twist doesn’t change my expectation of Palisade’s cost-effectiveness. I only slightly favor Center for AI Policy for the two reasons mentioned previously. I personally know people at Palisade, which I think biases me in its favor, and I might put Palisade at #3 if I wasn’t putting in mental effort to resist that bias.