A Policy Algebra for Trust-Preserving Agentic AI Execution
2026-08-17 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors explain that while AI agents often focus on being capable—like solving problems or using tools—this isn't enough to be trustworthy in important settings like businesses. They introduce the idea of "reliable capability," meaning agents must complete tasks following strict rules about identity, data access, budget, approvals, and audits. To manage this, they created a system called a policy algebra that helps control and combine these rules, making sure the agent behaves correctly even when multiple agents work together. Their approach was tested and found to stop most rule-breaking events while still completing most tasks, offering a way to build AI agents that are both smart and responsible.
large language modelagentic frameworkspolicy algebrareliable capabilitysecurity profilesruntime obligationsbudget constraintsaudit completenessmulti-agent systemsartifact materialization
Authors
Bhaskar Tripathi, Anurag Kumar, Ramendra Kumar, Bhavesh Gadhe
Abstract
Large language model-based agentic frameworks primarily optimize capability: whether an agent can reason, retrieve information, call tools, delegate work, and complete a goal. Enterprise execution requires a stronger property. A successful result is not reliable if it was produced through unauthorized data access, widened delegated authority, unapproved side effects, unrecoverable budget consumption, or incomplete evidence. This paper defines reliable capability as a path property: an agent is reliably capable only when it completes a task through action events that remain admissible under identity, profile, tool, data, memory, budget, artifact, approval, and audit constraints. We propose a policy algebra that defines the reliability envelope within which agent capability may be exercised. Security profiles and runtime obligations compose through joins, intersections, budget narrowing, approval inheritance, and evidence accumulation; the resulting composition is both trust-preserving and the least restrictive state satisfying all governing inputs. The algebra also propagates restrictions across multi-agent calls and introduces cost-aware artifact materialization, which redirects open-ended execution toward a recoverable outcome as budget exposure grows. The evaluation is interpreted as a reliability-capability trade-off rather than a capability benchmark: the policy-algebra runtime intervenes on 94.8% of policy-violating events while retaining an 86.9% task-completion rate, eliminates the observed profile-monotonicity and zero-artifact-exhaustion violations, and increases audit completeness to 98.6%. The method provides researchers and practitioners with formal correctness conditions, executable decision semantics, and trace evidence for building agents that are not only capable, but reliably capable.