We’re proposing an AI safety technique which trains agents to debate topics with one another, using a human to judge who wins.
We (along with researchers from Berkeley and Stanford) are co-authors on today’s paper led by Google Brain researchers, Concrete Problems in AI Safety. The paper explores many research problems around ensuring that modern machine learning systems operate as intended.
We’ve written a paper arguing that long-term AI safety research needs social scientists to ensure AI alignment algorithms succeed when actual humans are involved. Properly aligning advanced AI systems with human values requires resolving many uncertainties related to the psychology of human rationality, emotion, and biases.
arXiv:2606. 03812v1 Announce Type: new Abstract: Operational safety in high-stakes domains such as industrial process control, autonomous, and safety-critical systems, demand reliable hazard identification.
arXiv:2506. 13609v2 Announce Type: replace Abstract: Training powerful AI systems to exhibit desired behaviors hinges on the ability to provide accurate human supervision on increasingly complex tasks.
Ensuring that AI systems are built, deployed, and used safely is critical to our mission.
OpenAI introduces CoT-Control and finds reasoning models struggle to control their chains of thought, reinforcing monitorability as an AI safety safeguard.
arXiv:2510. 26518v2 Announce Type: replace Abstract: Human feedback is critical for aligning AI systems to human values.
In this work, we propose an agentic gamification framework for hazard-informed learning of robot safety policies through synthetic scenarios. We model scenario generation as an adversarial game between two agents: a Red Team that explores the space of potential failures by constructing hazardous situations, and a Blue Team that incrementally refines safety policies to prevent them.
One step towards building safe AI systems is to remove the need for humans to write goal functions, since using a simple proxy for a complex goal, or getting the complex goal a bit wrong, can lead to undesirable and even dangerous behavior. In collaboration with DeepMind’s safety team, we’ve developed an algorithm which can infer what humans want by being told which of two proposed behaviors is better.
RL-Teacher is an open-source implementation of our interface to train AIs via occasional human feedback rather than hand-crafted reward functions. The underlying technique was developed as a step towards safe AI systems, but also applies to reinforcement learning problems with rewards that are hard to specify.
We’re releasing Safety Gym, a suite of environments and tools for measuring progress towards reinforcement learning agents that respect safety constraints while training.
We’ve written a policy research paper identifying four strategies that can be used today to improve the likelihood of long-term industry cooperation on safety norms in AI: communicating risks and benefits, technical collaboration, increased transparency, and incentivizing standards. Our analysis shows that industry cooperation on safety will be instrumental in ensuring that AI systems are safe and beneficial, but competitive pressures could lead to a collective action problem, potentially causing AI companies to under-invest in safety.