Výsledky vyhledávání - "Williams, Zoe A"

Report

Mapping Technical Safety Research at AI Companies: A literature review and incentives analysis

Autor: Delaney, Oscar, Guest, Oliver, Williams, Zoe

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 DeepMind,

Externí odkaz: http://arxiv.org/abs/2409.07878

Zobrazit plný text záznamu

Report

Adapting cybersecurity frameworks to manage frontier AI risks: A defense-in-depth approach

Autor: Ee, Shaun, O'Brien, Joe, Williams, Zoe, El-Dakhakhni, Amanda, Aird, Michael, Lintz, Alex

The complex and evolving threat landscape of frontier AI development requires a multi-layered approach to risk management ("defense-in-depth"). By reviewing cybersecurity and AI frameworks, we outline three approaches that can help identify gaps in t

Externí odkaz: http://arxiv.org/abs/2408.07933

Zobrazit plný text záznamu

Report

Coordinated Disclosure of Dual-Use Capabilities: An Early Warning System for Advanced AI

Autor: O'Brien, Joe, Ee, Shaun, Kraprayoon, Jam, Anderson-Samways, Bill, Delaney, Oscar, Williams, Zoe

Advanced AI systems may be developed which exhibit capabilities that present significant risks to public safety or security. They may also exhibit capabilities that may be applied defensively in a wide set of domains, including (but not limited to) d

Externí odkaz: http://arxiv.org/abs/2407.01420

Zobrazit plný text záznamu

Report

Deployment Corrections: An incident response framework for frontier AI models

Autor: O'Brien, Joe, Ee, Shaun, Williams, Zoe

A comprehensive approach to addressing catastrophic risks from AI models should cover the full model lifecycle. This paper explores contingency plans for cases where pre-deployment risk management falls short: where either very dangerous models are d

Externí odkaz: http://arxiv.org/abs/2310.00328

Zobrazit plný text záznamu