What's New > CLAIRE GPT > Changed behavior
  

Changed behavior

The October 2025 release of CLAIRE GPT includes the following changed behavior.

Human intervention for clarifications

When CLAIRE GPT requires further information or detects ambiguities in a conversation, human intervention is possible to clarify such ambiguities. These human interventions make responses more conversational.
Previously, CLAIRE GPT responses didn't allow human intervention.

Proprietary third-party LLMs for inferences

CLAIRE generative AI services utilize proprietary LLMs from multiple third-party providers to perform specific functions. CLAIRE retrieves relevant metadata based on user requests and uses private and localized deployments of multiple LLMs for inferences.
Previously, LLMs in the CLAIRE global intelligence stack were used for inferences.
For more information, see Informatica Intelligent Data Management Cloud CLAIRE Security, Privacy, and Compliance Overview.