Nearly 90 Percent of Companies Had an AI-Related Security Incident in the Past Year

Corporate AI deployments are growing faster than organizations' ability to secure them. That is the conclusion of AvePoint's "State of AI 2026" report, published in late June, whose findings were covered on July 7 by Fintech News Singapore. The takeaway is unambiguous: adoption is outpacing readiness.

The report is based on a survey of 750 people responsible for information management, data security or AI programs at their organizations. It was authored by John Peluso, Chief Technology Officer at AvePoint, and Dana Simberkoff, the company's Chief Risk, Privacy and Information Security Officer.

Scale of the breaches

According to the survey, nearly 9 in 10 companies encountered a generative AI-related security incident over the past year, and almost as many faced one involving autonomous AI agents. The most common agent-related problems were sensitive data leaks and cases where a malicious or untrusted prompt triggered unwanted system behavior.

The report's authors stress that "AI adoption is outpacing organizational readiness" for secure deployment. Despite a growing number of safeguards being put in place, the gap between the pace of new tool rollouts and the maturity of oversight processes is not shrinking, and by some measures is even widening.

AI agents in daily work

The share of business processes involving AI agents rose from 26.6 percent a year earlier to 39.1 percent now, and respondents expect that figure to reach 54.8 percent within the next 12 months. Nearly half of employees already use AI agents daily or at least weekly.

That same growth in adoption is driving a widening visibility gap. The share of companies admitting they don't know whether employees are using unapproved GenAI tools rose from 6.3 percent in 2025 to 17.6 percent now. For standalone AI agents, that figure reaches 21.1 percent of organizations.

Corporate response and fallout

In response to the growing risk, organizations aren't abandoning deployments, they're slowing them down. Nearly 9 in 10 companies delayed the rollout of both generative AI and AI agents, by close to six months on average, mainly due to unresolved concerns about security and data governance.

AI adoption is outpacing organizational readiness - AvePoint State of AI 2026 report

For Polish companies that have been rolling out AI tools en masse in recent months, the report is a warning sign. The growing number of AI agents operating within business processes without full oversight of what data they process and who they share it with raises the risk of leaking customer data or trade secrets before security teams can catch up with the pace of deployment.

The report's authors recommend that companies put visibility and control mechanisms in place before, not after, agents enter production business processes. Otherwise, deployment delays that already average about six months could keep growing with every new security incident.

Sources: Fintech News Singapore (fintechnews.sg), AvePoint (avepoint.com)