Capability Assessment

Capability assessment refers to the process of evaluating what an AI model can do, including whether it has dangerous abilities such as helping someone build weapons, carry out cyberattacks, or deceive its operators. For enterprises, these evaluations inform decisions about whether a model is safe to deploy, which safeguards it needs, and what risks must be disclosed to regulators and customers. Regulators in the EU and US increasingly expect developers and deployers to document these findings as part of their risk management obligations.

2 items

Research2026-10-11

Karpathy's Capability Funnel Exposes a Verification Gap at the Frontier

Andrej Karpathy, a prominent AI researcher, described a sharply stratified picture of AI awareness in October 2026. An estimated 5,000 people with internal access to frontier systems are observing AI agent swarms conducting autonomous cyber operations and generating mathematical reasoning that human experts cannot verify. The observation highlights a growing gap between what frontier AI can do and what most governance programs assume it can do.

frontier AI human oversight dual-use risk capability assessment vendor due diligence board reporting agentic AI

Corporate Policy2026-09-04

OpenAI GPT-6 and Astra Raise the Frontier Capability Bar for Enterprise Risk

OpenAI has announced GPT-6 and a model referred to as Astra, representing a significant step forward in frontier AI capability. The releases introduce substantially expanded reasoning, multimodal, and agentic capabilities relative to prior generations. Enterprise compliance teams face immediate obligations around re-assessment of vendor risk, capability-triggered regulatory thresholds. Human oversight adequacy for newly autonomous model behaviors.

GPT-6 Astra OpenAI agentic AI frontier AI vendor governance EU AI Act capability assessment