Govern the whole AI estate, not just the use case in front of you.
Credo AI now governs your agents, MCP servers and deployment platforms alongside your use cases, models and vendors: one connected map of your AI estate, with risk that flows where it belongs and reviews that keep moving.
With Credo AI, you can now govern your whole AI estate, not just the use case in front of you. Agents, MCP servers and deployment platforms become first-class records in the Credo AI registry, linked to the use cases that depend on them. Risk you've already assessed flows to every use case that relies on it. Reviews move forward on their own once every reviewer has signed off, while people still make the decisions. And every step lands in the audit log.
Here is what's new, organized around the four things every governance program has to do: know what AI you run, reason about what applies, act with people in control, and prove it.
Know: one connected map of your AI estate
Agents, MCP servers and deployment platforms now join use cases, models and vendors as default entity types in Credo AI. Each one gets the same governance toolkit: its own stages, risk scoring, policy packs, questionnaires, triggers and actions, workflows, report templates and audit log.
The real value is in the links between them. Associate an agent with the use cases that rely on it, or a deployment platform with everything that runs on it, and the registry becomes a connected map instead of a flat list. When someone asks "which of our use cases run on this platform?" or "what depends on this MCP server?", you have the answer in one place.
What this means for you: a complete inventory of the AI you actually run, including the agentic layer that most programs can't see yet. If you track these assets somewhere else today, your Customer Success Manager can help you bring them in.
Reason: assess once, reuse everywhere
You shouldn't have to assess the same risk twice. With the new Associated Risk View, risk from upstream entities (a model, a vendor, an agent, a platform) carries through to every use case that depends on them, clearly labelled by source. Risk scoring now works for every entity type, and the right people are notified when a score changes.
Deciding what applies gets easier too. Scoping questionnaires work out which policy packs are relevant to a given system. Triggers and actions add the right follow-up questionnaires based on a user's answers. For example, a use case deployed in the EU automatically picks up the EU AI Act questionnaire.
What this means for you: less duplicated assessment work, and a clear line of sight from a vendor or model risk to every use case it touches.
Act: the platform moves the work, people make the decisions
Governance slows down when every step waits on an admin to click a button. Automatic stage and status transitions move a use case forward once every reviewer has approved. Stages that need a human decision stay manual, so you decide where automation helps and where sign-off matters.
Scheduled reviews now open on their own for every entity type. Webhooks and customizable email notifications tell the right people, and the right systems, what needs to happen next.
What this means for you: reviews that keep flowing without losing human oversight, and governance that connects to the tools your teams already use.
Prove: audit-ready by default
Evidence should build up as you govern, not get assembled in a scramble before an audit. Every entity type now carries a full audit log trail. Scheduled review flags make lapsed reviews visible before they become findings. And our standard Risk and Compliance Reports are now ready-made templates in Custom Reports, so you can start from them or build your own for any entity type.
What this means for you: evidence that's ready when a regulator, auditor or board asks for it.
All of this runs on a cleaner, more open API, with a single association endpoint and custom fields readable by name, so it's easier to wire Credo AI into your existing stack.
Regulatory coverage that stays current
This release adds 35 new policy packs, so your teams start from expert content instead of a blank page. New scoping questionnaires for healthcare and insurance then pick the packs that apply to each system.
What's new in the Industry:
Financial services
Financial Services Part 1: ready-made packs for banks, payment institutions and asset managers across the US (SR 26-2, NYDFS), the EU (DORA, MiFID II, PSD2) and Asia Pacific (APRA, MAS)
Healthcare
CHAI AI Intake, which your AI vendors can complete themselves, plus a healthcare scoping questionnaire.
Insurance
An insurance scoping questionnaire that selects the packs relevant to each system.
Cross-industry
Packs for California's AI statutes, and AIUC-1 updated to its July 2026 revision
Regulations don't stand still, and neither does our content. When SR 26-2 superseded SR 11-7 for model risk management, and when AIUC-1 published its July 2026 revision, we updated the packs so your program reflects the current rules without you rebuilding it.
Why a connected registry matters
Many tools can now list your agents or catalog your MCP servers. A list tells you what exists. It doesn't tell you which business use cases depend on an agent, what risk a platform passes on to everything running on it, or whether the right people have reviewed it this quarter.
That context is what governance runs on, and it's what Credo AI was built for. We're not a GRC tool adapted for AI. With this release, every agent, MCP server and deployment platform sits in the same connected registry as your use cases, models and vendors, under the same policies, the same risk methodology and the same audit trail. Your governance team, your model and third-party risk teams, and your engineers work from one system of record instead of stitching together several.
What's next
The October 28 release is a foundation. Each of these capabilities gets deeper from here, and we'll keep our policy packs current as standards like AIUC-1 continue to evolve.
We are building a platform that lets you know every AI system you run, reason about what applies, act with people in control, and prove it. It's the same platform Gartner named a Visionary in its inaugural Magic Quadrant for AI Governance Platforms, and that Forrester and IDC have each named a Leader.
Already a Credo AI customer? These capabilities are available now. Reach out to your Customer Success Manager to start registering your agents, MCP servers and deployment platforms, or see the full October 28 release notes.
New to Credo AI? Talk to an expert to see how Credo AI governs the agents you buy and the agents you build.
Read more: Pre-Release Notes: October 28, 2026 · Customer release notes
DISCLAIMER. The information we provide here is for informational purposes only and is not intended in any way to represent legal advice or a legal opinion that you can rely on. It is your sole responsibility to consult an attorney to resolve any legal issues related to this information.



