Enterprise AI Still Smarting From Leaping Before Looking
The Register, Tuesday, July 7th, 2026
A DigiCert survey finds 78% of firms hit AI security incidents amid a wide governance gap.
A DigiCert survey found nearly 78% of organizations suffered AI-related security incidents or vulnerabilities, driven mainly by unauthorized or misconfigured AI agents rather than bad code.
Although 90% discussed AI governance at board level, only 50% funded formal programs, and just 53% could trace AI decisions back to their models and source data. DigiCert argues AI agents need verified identities like employees, citing efforts such as Estonia's digital IDs and Microsoft's Agent ID.
The findings echo Spacelift research showing 93% of firms hit AI-caused infrastructure incidents, suggesting enterprises are prioritizing speed over risk assessment.