This page is still being polished. If you have thoughts, please share them via the feedback form.
Data on this page is preliminary and may change. Please do not share or cite these figures publicly.
Staged rollout strategies, phased deployment, and tiered access approaches for production systems.
Also in Operations & Security
Giving good faith, public interest evaluation researchers access to black-box research APIs that provide technical and legal safe harbours to limit barriers imposed by usage policy enforcement, logging, and stringent terms of service.
Many experts viewed this measure as generally beneficial for risk reduction across multiple areas. Several noted its potential to identify new risks and threat vectors, particularly for bias and discrimination, and chemical, biological, radiological, and nuclear (CBRN) risks. Multiple experts highlighted the value of independent scrutiny and collective intelligence in uncovering potential issues. However, concerns were raised by multiple experts about implementation challenges. These included determining who qualifies as a "good faith" researcher, managing potential conflicts of interest, and balancing transparency with information security risks. Several experts pointed out that API access alone may be limiting, suggesting full model access could be more effective. Some experts emphasised the measure's particular usefulness for bias and discrimination research, as it could allow minority groups to test models without significant financial barriers. Others noted it may be less effective for more diffuse risks like impacts on democratic processes. A minority of experts expressed scepticism about the measure's impact, citing limitations of black-box access or questioning whether it would significantly help with certain risk types like biosecurity.
Reasoning
Grants tiered API access to vetted researchers with relaxed usage policies, implementing staged deployment with controlled stakeholder access.
Pre-deployment risk assessments
Comprehensive risk assessments before deployment that would assess reasonably foreseeable misuse and include dangerous capability evaluations that incorporate post-training enhancements and collaborations with domain experts. Risk assessments would inform deployment decisions.
2.2.1 Risk AssessmentThird party pre-deployment model audits
External pre-deployment assessment to provide a judgment on the safety of a model. Auditors, which could be governments or independent third parties, would receive access to a fine-tuning API for testing, or further appropriate technical means.
2.2.3 Auditing & ComplianceExternal assessment of testing procedure
Bringing in external AI evaluation firms before deployment to assess and red-team the company's execution of dangerous capabilities evaluations.
2.2.2 Testing & EvaluationAdvanced model access for vetted external researchers
Examples of advanced access rights could include any of the following: increased control over sampling, access to fine-tuning functionality, the ability to inspect and modify model internals, access to training data, or additional features like stable model versions.
2.2.2 Testing & EvaluationData curation
Careful data curation prior to all development stages (including fine-tuning) to filter out high-risk content and ensure the training data is sufficiently high-quality.
1.1.1 Training DataHarmlessness training
State-of-the-art reinforcement learning and fine-tuning techniques, such as Reinforcement Learning from Human Feedback (RLHF) or Direct Preference Optimization (DPO), to ensure models do not engage in unsafe behavior.
1.1.2 Learning ObjectivesEffective Mitigations for Systemic Risks from General-Purpose AI
Uuk, Risto; Brouwer, Annemieke; Schreier, Tim; Dreksler, Noemi; Pulignano, Valeria; Bommasani, Rishi (2024)
The systemic risks posed by general-purpose AI models are a growing concern, yet the effectiveness of mitigations remains underexplored. Previous research has proposed frameworks for risk mitigation, but has left gaps in our understanding of the perceived effectiveness of measures for mitigating systemic risks. Our study addresses this gap by evaluating how experts perceive different mitigations that aim to reduce the systemic risks of general-purpose AI models. We surveyed 76 experts whose expertise spans AI safety; critical infrastructure; democratic processes; chemical, biological, radiological, and nuclear risks (CBRN); and discrimination and bias. Among 27 mitigations identified through a literature review, we find that a broad range of risk mitigation measures are perceived as effective in reducing various systemic risks and technically feasible by domain experts. In particular, three mitigation measures stand out: safety incident reports and security information sharing, third-party pre-deployment model audits, and pre-deployment risk assessments. These measures show both the highest expert agreement ratings (>60\%) across all four risk areas and are most frequently selected in experts' preferred combinations of measures (>40\%). The surveyed experts highlighted that external scrutiny, proactive evaluation and transparency are key principles for effective mitigation of systemic risks. We provide policy recommendations for implementing the most promising measures, incorporating the qualitative contributions from experts. These insights should inform regulatory frameworks and industry practices for mitigating the systemic risks associated with general-purpose AI.
Operate and Monitor
Running, maintaining, and monitoring the AI system post-deployment
Developer
Entity that creates, trains, or modifies the AI system
Govern
Policies, processes, and accountability structures for AI risk management
