Ethyca launches Astralis data-governance platform for real-time access control

Data-governance startup Ethyca launched Astralis on Aug. 4, a platform the company says cuts data-access approval times from about six weeks to five minutes using real-time, purpose-based access control.

Astralis enforces what Ethyca calls purpose-based access control – deciding whether a given use of data is permitted for its stated purpose – across data warehouses, notebooks, business-intelligence tools, AI agents and large language model workflows in real time.

The company’s headline claim is speed: shrinking data-access approvals from roughly six weeks to five minutes. Ethyca also says Astralis governs about 1 petabyte of data a day, makes some 150,000 policy decisions per second, and processes 744 million consent preferences a year, and that more than 200 global brands already use its platform.

Those figures come entirely from Ethyca and have not been independently verified; the six-weeks-to-five-minutes comparison, in particular, is a vendor benchmark with no disclosed methodology.

Astralis is organized in three layers: DETECT, built on Ethyca’s open-source Fides governance ontology; DECIDE, which the company calls a “Large Language Risk and Regulatory Model”; and ENFORCE, the access-control engine that applies the decisions. The pitch reflects a broader problem for companies wiring language models and AI agents into internal data, because access controls written for human users do not map cleanly onto software that can query everything at once.

Whether Astralis performs at the scale Ethyca describes, outside its own measurements, is the test that matters.