Build better policies based on FireTail’s topic-based guardrails.
As part of our latest release, we have made a change relating to topic-based guardrails and the policies you can build with FireTail.
.png)
What is changing.
As part of our September 2026 update, we’ve made a change. Now, you can only create new topic-based guardrail policies using FireTail's native topics and topic-based guardrails. You can no longer base a new policy on the topic-based guardrails available in AWS.
That is the whole change. If you do not build a new topic-based guardrail policy, you will notice nothing at all.
What this means for you.
Nothing, if you have policies in place today. Existing policies built on AWS topic-based guardrails continue to run exactly as they do now. There is no migration, no rebuild, and no action required.
The change applies only to new policy creation from this release onward.
Why we made this change.
FireTail's native topics and topic-based guardrails give you more room to define what a policy should catch, and they avoid some of the volume constraints that apply to the AWS approach. You can define topics and scale them on your terms.
Building new policies on FireTail's native capability also keeps your guardrails under one policy engine. Every topic-based guardrail lives alongside the rest of your AI governance policies, in one place, with one model for how they behave. As your AI footprint grows, that consistency is what keeps guardrails manageable.
.png)
The takeaway.
Guardrails should be simple to define and simple to scale. Native topics and topic-based guardrails make that the default for every new policy, without touching anything you already run.
Want to see topic-based guardrails at work across your AI estate? Schedule a Technical Deep Dive today
