GLACIS review
A hybrid runtime control layer for governing AI actions and recording decisions.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
GLACIS is a runtime governance and security platform for AI applications and agents. It is designed for teams that need policy controls at the point where AI actions occur, rather than only monitoring activity afterward. Its controls can allow, constrain, hold, or deny operations, while its hybrid deployment model supports a hosted workspace, self-hosted SDK, managed environments, and customer-environment deployment. The product is suited to organizations governing AI agents, tool calls, sensitive data, and externally directed actions.
Its standout capability is runtime decision-making across both inputs and outputs. GLACIS includes PII/PHI detection controls, prompt-injection and jailbreak detection, prohibited-use filtering, egress control, tool-call restrictions, and human-review routing. These controls give teams several response options instead of a simple pass-or-block workflow. Signed, hash-chained operational records and offline receipt verification add an evidence layer for governed decisions. The Explore plan includes 1,000 signed receipts per month, while production-focused plans add a dedicated arbiter, independent witnessing, evidence packs, retention, and production deployment. Enterprise adds private connectivity, customer-environment deployment, extended retention, named support, and an SLA.
GLACIS is a strong fit for teams seeking free access to hybrid runtime policy controls, with a 14-day trial also available. Its narrower ecosystem is an important consideration: documented SDK integrations cover OpenAI, Anthropic, and Google Gemini. Teams using other model providers or looking for a broadly open-source enforcement stack may need an alternative. GLACIS should appeal most to organizations that value explicit intervention decisions and verifiable records; teams needing only basic prompt filtering may find its governance model broader than necessary.
GLACIS pros and cons
- Where it wins
- Allow, constrain, hold, or deny AI actions at runtime
- Detects PII/PHI, prompt injection, and jailbreak attempts
- Signed, hash-chained records support offline verification
- Where it doesn't
- Integrations cover only OpenAI, Anthropic, and Google Gemini
- The runtime enforcement engine and Notary are not open source
- Advanced production capabilities are split across multiple plan listings
GLACIS fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update