Safety · Concept
Lean4Agent: Formal Verification of AI Agent Workflows
Framework using the Lean4 dependent-type formal language to model, verify, and debug AI agent workflows and execution trajectories for semantic consistency.
Forecast horizon
Connections
Connections · 3
How this node ties into the rest of the map, and the evidence behind each link.
to · constrains 3/10
Strategic Attack Selection in Agentic AI Control Evaluations Formal verification of agent workflows could constrain the attack surface exploited by strategic attack selection in agentic AI systems.
+3 growth from · supports 3/10
LLM Translation of Unstructured Requirements to LTL Automatic NL-to-LTL generation is a prerequisite step toward formal verification workflows for requirements.
+3 growth from · enables 3/10
LLM Translation of Unstructured Requirements to LTL Auto NL-to-LTL is a prerequisite step toward formal verification pipelines from requirements.
+3 growthSignal sources
Signal sources
Dated facts from primary sources in this direction.
US evaluation centre Jun 2025
In June 2025 the US AI Safety Institute was renamed the Center for AI Standards and Innovation (CAISI), pivoting toward security, standards and adversary-model assessment.
NIST → Frontier safeguards May 2025
Anthropic activated its ASL-3 deployment and security standard with Claude Opus 4 on 22 May 2025 — the first real-world trigger of a responsible-scaling tier, focused on blocking bio-weapon uplift.
Anthropic → Cross-border testing 2025
The International Network of AI Safety Institutes (launched Nov 2024) ran a third joint testing exercise focused on agentic AI systems across cyber and fraud strands.
European Commission — AI Office →