Open Ollama server detection

UpGuard’s latest update to Breach Risk and Vendor Risk improves external attack surface monitoring by detecting exposed Ollama instances—self-hosted large language models that become vulnerable when publicly accessible via port 11434. This new capability helps organizations prevent unauthorized access, safeguard AI assets, and mitigate the risk of data theft or model tampering.

Improvements to Trust Page questionnaires

It’s now easier to add security questionnaires to your Trust Page, with a redesigned module that streamlines the process and makes it easier to share your responses with prospects and customers. You can now import questionnaires directly on the Trust Page, as well as more easily access details about your Trust Page questionnaires via Answer Questionnaires > Public questionnaires.

Other improvements

• We’ve added company logos to Vendors in Vendor Risk.
• We’ve changed the handling of informational risks in the Vendor Risk Security profile. Details of informational risks are included but do not impact the control status or security rating.
• This release includes a number of bug fixes.