Comment by Jacob Charnock

We recommend that frontier AI developers begin hosting embedded assessments now, covering at least three areas central to managing risks from internal AI use: internal agent monitoring, internal agent security controls and permissions, and model alignment. [...] Evaluators should have parity of access with internal staff conducting similar risk assessments or senior alignment researchers, as appropriate. This should include relevant infrastructure, personnel, and documentation, scoped to the assessment targets. Access to relevant information should be granted by default.