AI Safety

AI safety research focuses on building AI systems that are safe, beneficial, and aligned with human values. Key concerns include preventing misuse, ensuring robustness and interpretability, and governing systems that become increasingly capable. This includes both technical research (alignment techniques, interpretability, adversarial robustness) and institutional approaches (governance, standards, oversight).

Key papers

• Truthful AI: Developing and Governing AI That Does Not Lie — governance and technical approaches to truthful AI as a component of AI safety
• Mirsky et al. (2021) — The Threat of Offensive AI to Organizations — threat modeling for AI-enabled attacks

Related topics

• AI governance and compliance (institutional mechanisms)
• Truthfulness (preventing false claims as a safety concern)
• Content moderation (governance of harmful content)