Artificial intelligence has become a core consideration for cybersecurity teams, but many organizations remain reluctant to trust AI with autonomous security decisions, according to Arctic Wolf's newly released 2026 AI & Cybersecurity Trends Report.

The report, based on a global survey of 1,350 security and IT decision-makers conducted by Sapio Research, found that 94% of organizations now use large language models and an equal percentage say AI capabilities influence cybersecurity purchasing decisions. More than half of respondents, 51%, said AI functionality is now a requirement when evaluating cybersecurity vendors.

Despite widespread adoption, confidence in AI-driven security operations remains limited. Only 14% of organizations said AI is central to their security operations strategy, while just 53% trust AI to perform narrowly defined security actions such as blocking malicious IP addresses at a firewall due to concerns about false positives.

The findings come as organizations continue to face significant cyber risks. Nearly two-thirds, or 63%, of respondents said they experienced a significant cybersecurity incident during the past year, with another 7% unsure whether they had experienced such an event. Among organizations affected by an incident, 48% reported disruptions lasting two weeks or longer.

At the same time, 96% of security leaders said they remain confident their teams can keep pace with the volume and complexity of modern threats, highlighting what the report describes as a disconnect between perceived preparedness and real-world outcomes.

The report also found that AI has become both a strategic priority and a leading concern. Thirty-five percent of respondents identified AI as their top cybersecurity risk for the second consecutive year, surpassing ransomware and malware, even as ransomware remains the most common incident response case handled by Arctic Wolf Incident Response.

While organizations expressed optimism about AI's potential, trust continues to be a major obstacle to broader adoption. Eighty-five percent of respondents believe AI will improve their ability to detect new or difficult-to-identify threats, and 72% believe AI is more capable than humans at identifying threats. However, concerns about data privacy, cited by 51% of respondents, lack of human intuition at 49%, and questions around accountability and inaccurate outcomes continue to slow adoption of agentic AI technologies.

The report also examined the operational challenges facing security teams. Respondents said they spend roughly 13 to 15 hours each week across major security functions including reducing false positives, managing security tools and meeting compliance requirements, underscoring continued demand for automation and AI.

Data loss emerged as the most damaging type of cyber incident reported. Twenty-one percent of respondents said inadvertent data exposure or accidental data loss was the most significant cybersecurity incident they experienced, while another 8% identified deliberate data exfiltration as a serious concern.

Ransomware also continues to affect organizations, particularly in North America, where 74% of organizations that experienced a ransomware attack ultimately made a payment directly or through a third party, the highest rate among the regions surveyed.

"The challenge now is trust," said Adam Marrè, Chief Information Security Officer at Arctic Wolf. "Security leaders want the speed, scale, and efficiency AI can deliver, but they also need confidence that the outcomes are trustworthy, explainable, and aligned to their business. The future of security operations will belong to organizations that combine trusted AI with human expertise to move faster, make better decisions, and build greater resilience."