AI Governance

TL;DR for operators An AI agent can authenticate successfully while its credential, software client, or external identity record is associated with the wrong persistent actor. This becomes consequential when agents operate across deployments, rotate credentials, and accumulate audit histories. SoulAuth proposes a stable ActorIdentity for each independently identifiable Human or long-lived AI actor. Credentials, clients, sessions, and cross-domain identity relationships remain separate objects with their own lifecycles. Authentication proves an identity under specified conditions; it does not automatically establish permission to act. ...

TL;DR for operators An LLM judge can agree closely with human reviewers, produce consistent scores, and still provide insufficient evidence for a product decision. Agreement measures correspondence between evaluation methods; it does not establish that either method adequately captures the quality or risk the decision requires. LLJ Cards proposes five documentation and evaluation components, containing 45 guiding questions, to address this gap. The framework makes evaluation context, measurement criteria, prototyping, pipeline specifications, and validity evidence explicit. ...

TL;DR for operators An AI compliance platform can execute technical tests correctly without resolving whether the resulting evidence addresses the right regulatory questions. Engineers, assessment specialists, and regulators bring different responsibilities and interpretations to the same system. A new study develops a collaborative assessment tool that combines AI system qualification, regulatory controls, technical tests, shared dashboards, and reporting. Three design iterations revealed why testing functionality alone was insufficient: some stakeholders struggled to interpret the original technical interface, while others needed assessment pathways that did not begin with executable tests. ...

TL;DR for operators An AI monitor that performs well before an agent begins learning may become unreliable as the agent adapts to its judgments. Monitor accuracy is therefore an ongoing property of the worker-monitor interaction, not merely a deployment-time benchmark. Research on continuously co-trained monitors establishes a theoretical condition under which occasional reliable corrections can support increasingly accurate oversight while the fraction of externally verified actions approaches zero. When legitimate work also remains sufficiently rewarding, better monitoring can discourage cheating rather than merely detect it. ...

TL;DR for operators Enterprise AI approval should attach to the complete workflow configuration, not to a model name or benchmark score. EnterpriseVal formalizes that idea by freezing the model, prompts, retrieval or memory, tools, guardrails, logging, and human-review design before evaluation. The framework then tests six families of evidence—fidelity, utility, efficiency, reliability, assurance, and oversight—and uses precommitted thresholds with uncertainty to classify a workflow as Reject, Conditional, or Scale. Only candidates that survive those risk gates should be ranked, and the preferred candidate can change when the business changes the relative importance of its metrics. ...

TL;DR for operators A higher-priced model can be the cheaper model to run. Under the September 2026 pricing assumptions examined by Abbasi, Aqrawi, and Kwartler, Fable 5.1 becomes cheaper than Opus 5 once a warm agentic turn reaches roughly 71,000 tokens of context at 18 tool steps, despite Fable’s higher ordinary input and output prices.1 ...

TL;DR for operators A fixed red-team suite answers a predetermined set of questions. CART asks an additional operational question: given what the last tests revealed, where should the next test budget go? In an illustrative 99-round run, the first 12 reconnaissance tests produced zero scores. A cascading-hallucination signal appeared at round 13. From there, 67% of the remaining probes were directed toward the two most brittle categories, while testing strategies continued to vary; the peak risk score reached 4.37. ...

TL;DR for operators A clinical LLM can look strong on benchmark accuracy and still behave poorly when the evaluation asks a different question: will a small clinically meaningful change move its answer in the intended direction, and will clinically unchanged evidence remain stable when demographic descriptors are added? The paper tests those properties separately rather than compressing them into one safety score. Across six models and 150 MedQA questions, overall counterfactual validity was only 25.3%. In a separate demographic stress test, 20.3% of 5,400 baseline-to-prefix comparisons changed the selected answer. Explanation drift and automated stereotype flags added two further screening dimensions. ...

TL;DR for operators A pricing agent can receive the same underlying market facts and still behave differently because those facts arrive in a different format, order, or textual context. Lee and Park call this market signal injection: presentation-level manipulation of LLM pricing inputs without an explicit instruction to set a particular price.1 Their simulations show that the effect can extend beyond the manipulated agent. Under undefended sentiment-context attacks, non-target firms’ mean price shifts were about 83–99% of the target firm’s shift in duopoly and 65–99% in triopoly. ...

TL;DR for operators A policy-aware LLM has to do more than retrieve the right document. It must know what evidence is mandatory, when evidence is insufficient, when a case requires human review, which policy version governed the decision, and what information must be retained so the decision can later be reconstructed. Policy-as-Skill: Governed LLM Decision Support with Evidence, Deterministic Control, and Audit1 turns those obligations into explicit runtime components rather than leaving them inside prompts or application glue code. ...