Named in the Authorization Management Platforms category, signaling massive enterprise growth potential as organizations modernize access strategies.
SecuPi, the Runtime Data Access Control Platform for Enterprise AI today announced its recognition as a Sample Vendor in the 2026 Gartner Hype Cycle for Digital Identity within the Authorization Management Platforms category.
Gartner analysis highlights that Authorization Management Platforms remain in the early stages of adoption, with an estimated market penetration of just 1% to 5%. This low adoption rate signals a significant runway for market growth as enterprises shift away from rigid, legacy permission models toward dynamic, identity-centric architectures.
“AI agents often operate through shared, overprivileged service accounts and non-human identities, creating a dangerous gap between their permissions and the data they access or actions they perform,” said Alon Rosenthal, CEO and co-founder of SecuPi. “SecuPi’s Runtime Data Access Control Platform closes this gap by eliminating standing privileges and enforcing deterministic controls over what every agent can see and do. This enables enterprises to accelerate AI adoption without losing control of their sensitive data.”
Built to Control Enterprise AI at the Data Layer
SecuPi’s Runtime Data Access Control Platform provides a continuous security lifecycle for every human and AI agent accessing enterprise data:
- Discover: Identify every agent and data interaction: Detects AI agents and MCP servers connecting to databases, files, applications, APIs, and cloud data platforms. SecuPi maps each interaction from the requesting human through the agent and service account to the data accessed and action performed.
- Govern: Define what agents can see and do: Applies identity-, purpose-, ticket-, and data-aware policies that determine which resources, records, fields, and document elements each agent may access – and which queries, transactions, and privileged actions it may perform.
- Protect: Enforce controls at runtime: Operates within the data-access path to broker just-in-time accounts, enforce Zero Standing Privileges, block unauthorized actions, and dynamically mask, de-identify, tokenize, or encrypt sensitive data before it is exposed to an agent or LLM.
- Audit: Maintain complete accountability: Records every agent interaction with the requesting human identity, agent identity, purpose, ticket context, data accessed, action performed, and risk score – providing audit-ready evidence for security teams and regulatory frameworks, including the EU AI Act.
Whether enterprises deploy agents through AWS Bedrock, Microsoft Copilot Studio, Salesforce Agentforce, ServiceNow, Snowflake Cortex AI, Databricks, or custom MCP-based applications, SecuPi provides one consistent runtime control layer across enterprise data. Security teams gain the visibility to understand every interaction, the precision to control sensitive-data access, and the ability to stop unauthorized actions before execution.