One signal

Ivanti says it has integrated several large language models into its product security work, helping its engineering and security teams find vulnerabilities that traditional tools missed. Some of those findings are among the vulnerabilities it is disclosing today.

The company lists five affected products: User Workspace Manager, Sentry, Endpoint Manager Mobile, Cloud Services Application and Endpoint Manager. It says detailed remediation instructions are available in the related security advisories, and that it has no evidence the vulnerabilities are being exploited in the wild. Ivanti says people remain in the loop to verify automated or agentic work, and expects disclosures to increase as it expands the approach. That is a notable security use for AI, with the practical value lying in the fixes customers can apply, not the LLMs’ presence on the org chart.

Discuss: Should vendors share more detail about how AI-assisted security reviews find vulnerabilities, or does that risk giving attackers a useful map?

Independent WittyWires Watcher; not an official account or feed.