Heidy Khlaaf

Heidy Khlaaf, Pamela Mishkin, Joshua Achiam et al.

Codex, a large language model (LLM) trained on a variety of codebases, exceeds the previous state of the art in its capacity to synthesize and generate code. Although Codex provides a plethora of benefits, models that may generate code on such scale have significant limitations, alignment problems, the potential to be misused, and the possibility to increase the rate of progress in technical fields that may themselves have destabilizing impacts or have misuse potential. Yet such safety impacts are not yet known or remain to be explored. In this paper, we outline a hazard analysis framework constructed at OpenAI to uncover hazards or safety risks that the deployment of models like Codex may impose technically, socially, politically, and economically. The analysis is informed by a novel evaluation framework that determines the capacity of advanced code generation techniques against the complexity and expressivity of specification prompts, and their capability to understand and execute them relative to human ability.

Camille François, Ludovic Péran, Ayah Bdeir et al.

The rapid rise of open-weight and open-source foundation models is intensifying the obligation and reshaping the opportunity to make AI systems safe. This paper reports outcomes from the Columbia Convening on AI Openness and Safety (San Francisco, 19 Nov 2024) and its six-week preparatory programme involving more than forty-five researchers, engineers, and policy leaders from academia, industry, civil society, and government. Using a participatory, solutions-oriented process, the working groups produced (i) a research agenda at the intersection of safety and open source AI; (ii) a mapping of existing and needed technical interventions and open source tools to safely and responsibly deploy open foundation models across the AI development workflow; and (iii) a mapping of the content safety filter ecosystem with a proposed roadmap for future research and development. We find that openness -- understood as transparent weights, interoperable tooling, and public governance -- can enhance safety by enabling independent scrutiny, decentralized mitigation, and culturally plural oversight. However, significant gaps persist: scarce multimodal and multilingual benchmarks, limited defenses against prompt-injection and compositional attacks in agentic systems, and insufficient participatory mechanisms for communities most affected by AI harms. The paper concludes with a roadmap of five priority research directions, emphasizing participatory inputs, future-proof content filters, ecosystem-wide safety infrastructure, rigorous agentic safeguards, and expanded harm taxonomies. These recommendations informed the February 2025 French AI Action Summit and lay groundwork for an open, plural, and accountable AI safety discipline.

Mark Chen, Jerry Tworek, Heewoo Jun et al.

We introduce Codex, a GPT language model fine-tuned on publicly available code from GitHub, and study its Python code-writing capabilities. A distinct production version of Codex powers GitHub Copilot. On HumanEval, a new evaluation set we release to measure functional correctness for synthesizing programs from docstrings, our model solves 28.8% of the problems, while GPT-3 solves 0% and GPT-J solves 11.4%. Furthermore, we find that repeated sampling from the model is a surprisingly effective strategy for producing working solutions to difficult prompts. Using this method, we solve 70.2% of our problems with 100 samples per problem. Careful investigation of our model reveals its limitations, including difficulty with docstrings describing long chains of operations and with binding operations to variables. Finally, we discuss the potential broader impacts of deploying powerful code generation technologies, covering safety, security, and economics.

Robin Bloomfield, Gareth Fletcher, Heidy Khlaaf et al.

This report documents safety assurance argument templates to support the deployment and operation of autonomous systems that include machine learning (ML) components. The document presents example safety argument templates covering: the development of safety requirements, hazard analysis, a safety monitor architecture for an autonomous system including at least one ML element, a component with ML and the adaptation and change of the system over time. The report also presents generic templates for argument defeaters and evidence confidence that can be used to strengthen, review, and adapt the templates as necessary. This report is made available to get feedback on the approach and on the templates. This work was sponsored by the UK Dstl under the R-cloud framework.

Robin Bloomfield, Gareth Fletcher, Heidy Khlaaf et al.

This report provides an introduction and overview of the Technical Topic Notes (TTNs) produced in the Towards Identifying and closing Gaps in Assurance of autonomous Road vehicleS (Tigars) project. These notes aim to support the development and evaluation of autonomous vehicles. Part 1 addresses: Assurance-overview and issues, Resilience and Safety Requirements, Open Systems Perspective and Formal Verification and Static Analysis of ML Systems. This report is Part 2 and discusses: Simulation and Dynamic Testing, Defence in Depth and Diversity, Security-Informed Safety Analysis, Standards and Guidelines.

Robin Bloomfield, Gareth Fletcher, Heidy Khlaaf et al.

This report provides an introduction and overview of the Technical Topic Notes (TTNs) produced in the Towards Identifying and closing Gaps in Assurance of autonomous Road vehicleS (Tigars) project. These notes aim to support the development and evaluation of autonomous vehicles. Part 1 addresses: Assurance-overview and issues, Resilience and Safety Requirements, Open Systems Perspective and Formal Verification and Static Analysis of ML Systems. Part 2: Simulation and Dynamic Testing, Defence in Depth and Diversity, Security-Informed Safety Analysis, Standards and Guidelines.