TenetGraph provides runtime authorization for AI agents, deriving precise permissions for each action and enforcing rule‑based controls on data access, financial transactions, and system operations. By generating auditable decision records that prove compliance with policies such as “no PII sharing,” it helps organizations prevent accidental or malicious harm from agents that have broad access to corporate resources.
Funding
Funding not disclosed
Founders
Product
Problem
AI agents often require extensive permissions to access CRM, email, and payment systems, giving them broad standing access that can unintentionally expose sensitive data such as personally identifiable information (PII). Without runtime controls, agents may perform actions that exceed their intended scope, leading to data leakage or other harmful outcomes.
Solution
TenetGraph provides a runtime authorization layer that evaluates each action an AI agent attempts against a defined policy set. The platform derives the specific permissions for each agent, generates an auditable decision record for every request, and enforces allow/deny outcomes in real time. By tying decisions to explicit rules (e.g., prohibiting PII sharing), TenetGraph ensures that only authorized data is returned while all actions are logged for compliance and audit purposes. The system integrates with existing AI workflows, allowing developers to define granular policies without modifying the agent’s core logic.
Target Audience
Primary customers are enterprises deploying AI agents for customer support, sales, or finance operations that need to protect sensitive data and meet compliance requirements.
Features
- Real-time policy evaluation for every agent action, preventing out‑of‑scope data access
- Automatic generation of immutable decision logs that detail the rule applied and the outcome
- Fine‑grained permission modeling that can restrict specific fields (e.g., PII) while allowing other data
- Simple policy authoring language to define rules such as “pii.share = false”
- Integration hooks for common AI platforms, enabling enforcement without code changes
- Auditable reports that provide regulators and auditors with clear evidence of compliance