Semperis Study Warns of Rising AI Risks to Identity Systems

Semperis announced in a press release the results of a study examining the impact of artificial intelligence on identity system security. The research, based on responses from 1,100 organizations across eight countries, found that 74% of companies believe AI will increase attacks on identity infrastructure such as Active Directory, Entra ID, and Okta.

The study shows that 93% of organizations already use or plan to use AI agents for sensitive tasks including password resets and VPN access. Despite this, only 32% of respondents said they are very confident they could regain control if AI exposed administrator credentials. In the United States, 53% expressed confidence in recovery, compared to just 12% in France.

According to the report, 92% of respondents said AI is installed on local machines with access to SSH and encryption keys, while only 65% have AI identities fully registered, authenticated, and authorized. Six percent admitted they do not track AI identities at all. Among those that do, 57% use the same system as for human identities and 43% rely on separate systems.

The study also found that 29% of organizations already use AI agents to manage security help desk tickets, and another 65% plan to do so within a year. On a positive note, 83% of respondents said AI identity governance is a priority for the coming months. The findings highlight the growing need for clear identity governance practices and recovery readiness as AI becomes embedded in enterprise operations.