32% of AI/LLM findings are rated as high risk, nearly 2.7x the overall high-risk rate of 12%.
32% of AI/LLM findings are rated as high risk, nearly 2.7x the overall high-risk rate of 12%. — This cybersecurity statistic was published by Cobalt in May 2026. It covers topics including AI Testing, LLM Testing, Vulnerabilities. The original data appears in State of Pentesting Report 2026. For the full methodology and detailed findings, refer to the original report.
Share or Copy this stat
Frequently Asked Questions
What does this statistic say?
32% of AI/LLM findings are rated as high risk, nearly 2.7x the overall high-risk rate of 12%. This data was published by Cobalt and covers AI Testing, LLM Testing, Vulnerabilities.
Where does this data come from?
This statistic comes from State of Pentesting Report 2026, published by Cobalt on May 27, 2026. You can view the original report at https://resource.cobalt.io/state-of-pentesting-2026?.
What cybersecurity topics does this cover?
This statistic relates to AI Testing, LLM Testing, Vulnerabilities. Browse more statistics on AI Testing or from Cobalt.