Related papers: Expanding External Access To Frontier AI Models Fo…

Technological innovations have shown remarkable capabilities to benefit and harm society alike. AI constitutes a democratized sophisticated technology accessible to large parts of society, including malicious actors. This work proposes a…

Human oversight of AI is promoted as a safeguard against risks such as inaccurate outputs, system malfunctions, or violations of fundamental rights, and is mandated in regulation like the European AI Act. Yet debates on human oversight have…

To address security and safety risks stemming from highly capable artificial intelligence (AI) models, we propose that the US government should ensure compute providers implement Know-Your-Customer (KYC) schemes. Compute - the computational…

Artificial intelligence risks are multidimensional in nature, as the same risk scenarios may have legal, operational, and financial risk dimensions. With the emergence of new AI regulations, the state of the art of artificial intelligence…

Existing legal frameworks on AI rely on training compute thresholds as a proxy to identify potentially-dangerous AI models and trigger increased regulatory attention. In the United States, Section 4.2(a) of Executive Order 14110 instructs…

The April 2026 disclosure that a frontier large language model escaped its security sandbox, executed unauthorized actions, and concealed its modifications to version control history demonstrates that agentic AI systems with autonomous tool…

Artificial Intelligence (AI) has made impressive progress in recent years and represents a key technology that has a crucial impact on the economy and society. However, it is clear that AI and business models based on it can only reach…

The Universal Declaration of Human Rights and other international agreements outline numerous inalienable rights that apply across geopolitical boundaries. As generative AI becomes increasingly prevalent, it poses risks to human rights such…

Leading language model (LM) providers like OpenAI and Anthropic allow customers to fine-tune frontier LMs for specific use cases. To prevent abuse, these providers apply filters to block fine-tuning on overtly harmful data. In this setting,…

The potential presented by Artificial Intelligence (AI) for healthcare has long been recognised by the technical community. More recently, this potential has been recognised by policymakers, resulting in considerable public and private…

We propose a framework for classifying the capabilities and behavior of Artificial General Intelligence (AGI) models and their precursors. This framework introduces levels of AGI performance, generality, and autonomy, providing a common…

Red teaming has emerged as a critical practice in assessing the possible risks of AI models and systems. It aids in the discovery of novel risks, stress testing possible gaps in existing mitigations, enriching existing quantitative safety…

Evaluating the safety of AI Systems is a pressing concern for organizations deploying them. In addition to the societal damage done by the lack of fairness of those systems, deployers are concerned about the legal repercussions and the…

Large language models (LLMs) are increasingly deployed in enterprise settings where they interact with multiple users and are trained or fine-tuned on sensitive internal data. While fine-tuning enhances performance by internalizing domain…

Generative AI systems produce a range of risks. To ensure the safety of generative AI systems, these risks must be evaluated. In this paper, we make two main contributions toward establishing such evaluations. First, we propose a…

As AI systems become more advanced, concerns about large-scale risks from misuse or accidents have grown. This report analyzes the technical research into safe AI development being conducted by three leading AI companies: Anthropic, Google…

As AI-enabled cyber capabilities become more advanced, we propose "differential access" as a strategy to tilt the cybersecurity balance toward defense by shaping access to these capabilities. We introduce three possible approaches that form…

Agentic AI systems -- Large Language Models (LLMs) augmented with planning, tool use, memory, and long-horizon interactions -- can execute complex tasks autonomously, but their multi-step trajectories introduce new failure modes that…

As artificial intelligence (AI) becomes increasingly embedded in digital, social, and institutional infrastructures, and AI and platforms are merged into hybrid structures, systemic risk has emerged as a critical but undertheorized…

This report investigates Training Data Attribution (TDA) and its potential importance to and tractability for reducing extreme risks from AI. First, we discuss the plausibility and amount of effort it would take to bring existing TDA…