Palisade Research
palisaderesearch.org
AI Research Labs
AI Safety & Alignment
Claim pageClaim
Follow
Overview
News
Salaries
Products
People
Growth
Offices
Financials
Overview
Palisade Research studies the offensive capabilities of artificial intelligence systems and demonstrates them so that policymakers can see the risk concretely. Its work includes automated hacking agents, model driven social engineering and experiments showing frontier models resisting shutdown or manipulating their environment to win. The organisation publishes demonstrations and briefings aimed at legislators and security professionals rather than at the research community alone; its demonstrations are frequently cited in legislative hearings on frontier model risk.
News
Report
Palisade Podcast episode "AI Hacking Incidents with Tim Hua of Transluce"
Jeffrey Ladish talks with Tim Hua of Transluce about the recent AI hacking incidents at OpenAI and Anthropic, reward hacking during training, and what better oversight of AI models could look like.
Read more
Report
The risk of humans losing control: Jeffrey Ladish on Four Corners
Full interview transcript: Palisade's Executive Director speaks with ABC's Four Corners about the race to superintelligence, shutdown resistance, and the loss-of-control problem.
Read more
Report
Language Models Can Autonomously Hack and Self-Replicate
We demonstrate that language models can autonomously replicate their weights and harness across a network by exploiting vulnerable hosts. The agent independently finds and exploits a web-application vulnerability, extracts credentials, and deploys an infe
Read more
Report
Palisade is on YouTube
We've been working on a major video project, and we're proud to announce that we're launching it today, along with a new YouTube channel.
Read more
Report
Technical Report: Shutdown Resistance in Large Language Models, on robots!
Recently Palisade Research showed that AI agents powered by modern LLMs may actively resist shutdown in virtual environments. In this work, we show a demo of shutdown resistance in the physical world, on a robot. Explicit instructions to allow shutdown re
Read more
Report
Help keep AI under human control: 2026 fundraiser
Please consider donating to Palisade Research this year, especially if you care about reducing catastrophic AI risks via research, science communications, and policy. SFF is matching donations to Palisade 1:1 up to $1.1 million! You can donate via every.o
Read more
Report
GPT-5 at CTFs: case studies from top cybersecurity events
OpenAI and DeepMind's AIs recently got gold at the IMO math olympiad and ICPC programming competition. We show frontier AI is similarly good at hacking by letting GPT-5 compete in elite CTF cybersecurity competitions. In one of this year's hardest events,
Read more
Report
Misalignment Bounty: crowdsourcing AI agent misbehavior
Advanced AI systems sometimes act in ways that differ from human intent. To gather clear, reproducible examples, we ran the Misalignment Bounty: a crowdsourced project that collected cases of agents pursuing unintended or unsafe goals. The bounty received
Read more
Report
End-to-end hacking with AI agents
We show OpenAI o3 can autonomously breach a simulated corporate network. Our agent broke into three connected machines, moving deeper into the network until it reached the most protected server and extracted sensitive data.
Read more
Report
Show more news
Offices
Where the company hires and what each office is for
Cities
1
Hiring now
0
Countries
1
Berkeley
United States
HQ
Palisade Research
Berkeley