Darktrace Survey Finds Formal AI Policies Declined to 37%

Darktrace detailed in a company report that 37% of organizations have a formal policy for secure AI use, down from 45% a year earlier. The findings come from a survey of more than 1,500 cybersecurity professionals across 14 countries.

Generative AI is present in 77% of security technology stacks, but 74% of respondents limit its autonomy in security operations centers. Only 14% allow AI to act without human involvement.

The survey found that 46% lack confidence in their ability to defend against attacks powered by AI. Another 73% said these threats already have a significant impact on their organizations, while 92% said they are forcing major defensive upgrades.

CISOs reported feeling the impact of AI threats at a higher rate than security operations staff, at 78% compared with 68%. Threat detection and anomaly identification ranked as the areas where respondents expect defensive AI to have the greatest impact.