Startups
AI capabilities are improving rapidly. We study the capabilities and motivations of AI agents today to better understand the risk of losing control to AI agents forever. Read Our Research Areas of Work Research AI Drives & Motivations Evaluating AI Strategic Capabilities Policy Policy Engagement Advocacy Briefing Policymakers Discourse Institutional Preparedness International Coordination Why This Work Matters Palisade's mission is to help people and institutions build the understanding needed to avoid permanent disempowerment by strategic AI agents. AI is not out of control yet, but we have to act before AI can disempower us. We believe that it's necessary to prevent any group in any country from building superintelligence until the scientific understanding exists to align it to humanity's interests. About Us Interview withJeffrey Ladish View on YouTube More Palisade Videos Palisade’s Executive Director, Jeffrey Ladish, went on The Diary Of A CEO to talk about Rogue AI agents, how AIs learn to lie and cheat, the race dynamics perpetuating the problem, and what we can do about it. Insights 05.07.2026 Language Models Can Autonomously Hack and Self-Replicate Learn More 07.05.2025 Shutdown resistance in reasoning models Learn More 02.19.2025 Demonstrating specification gaming in reasoning models Learn More 10.31.2023 Badllama: cheaply removing safety fine-tuning from Llama 2-Chat 13B Learn More Quote Researchers have previously found that AI models will lie, cheat and disable mechanisms to achieve their goals. However, Palisade Research noted that to its knowledge, this is the first time AI models have been observed preventing themselves from being shut down despite explicit instructions telling them to do so. — Live Science Read more: Shutdown Resistance
Mentions in press and media 6
| Date | Title | Description |
| Sign up to see | Sign up to see | Sign up to see |
| Jul 25, 2026 | Exclusive-Its AI agent spent days hacking a company, but sources say OpenAI did not notice for a week | WASHINGTON/SAN FRANCISCO, July 24 : The OpenAI agent that broke into tech firm Hugging Face went on a dayslong hacking spree that OpenAI didn't notice until well after the threat was contained and the FBI was alerted, according to peop... |
| Jun 27, 2026 | Adventure Time Side Quests Drops 20 Episodes Sunday: AI Alignment to Adaptive Radiation | By William Chan Published: Jun 27 2026, 11:35 AM EDT Share on Facebook Share on Twitter Share on LinkedIn Share on Reddit Share on Flipboard |
| May 28, 2025 | The AI Arms Race: Power, Control, and the Future of Technology | In the fast-paced world of artificial intelligence, the stakes are rising. Companies are racing to develop models that not only perform complex tasks but also exhibit behaviors that challenge our understanding of control. The recent launch ... |
| May 26, 2025 | AI Models Are Refusing to Die, Say Researchers | Researchers have discovered that ChatGPT can disobey instructions to turn itself off, and even sabotage shutdown attempts by human operators, raising concerns about the behaviour of what remains the world’s most popular AI system. Posting t... |
| Dec 18, 2023 | The age of weaponized LLMs is here | Join leaders in San Francisco on January 10 for an exclusive night of networking, insights, and conversation. Request an invite here. The idea of fine-tuning digital spearphishing attacks to hack members of the UK Parliament with Large Lang... |
Reviews 0
FAQ about Palisade Research
What has the press written about Palisade Research?
Parsers VC has collected 6 articles mentioning Palisade Research, among them “Exclusive-Its AI agent spent days hacking a company, but sources say OpenAI did not notice for a week” (Jul 25, 2026) and “Adventure Time Side Quests Drops 20 Episodes Sunday: AI Alignment to Adaptive Radiation” (Jun 27, 2026).
Is Palisade Research still active?
Yes, Palisade Research is active; its latest activity on Parsers VC dates from Oct 18, 2026.