IBM said July 29 that 85% of organizations in a May follow-on study planned to increase security spending after becoming aware of advanced frontier-model cyber capabilities, compared with 64% in its broader breach study who planned increases after a breach. Three quarters said the threats were changing how they deploy agents across security operations; more than half used agents for detection and containment, but 18% used them for vulnerability management.
IBM said in a July 29 release that 85% of organizations in follow-on research planned to increase security spending after becoming aware of advanced frontier AI cyber capabilities. That compared with 64% in IBM's broader breach research who said they planned to increase security spending after experiencing a breach.
The 2026 Cost of a Data Breach Report covered breaches experienced by 602 organizations worldwide from March 2025 through February 2026. A May follow-on study received responses from 456 of those organizations; IBM said 78%, or 356 organizations, were aware of recent reports about highly advanced frontier models. These are survey responses and spending intentions, not a forecast of total market spending.
Agent deployment is uneven
IBM reported that three quarters of organizations said frontier AI threats were prompting them to rethink how agents are deployed across security operations. More than half reported using agents for threat detection and containment, while 18% used them for vulnerability management.
That gap matters because faster vulnerability discovery does not reduce exposure unless teams can validate findings, prioritize fixes, and complete remediation. IBM framed the interval between discovery and remediation as a central operational risk as AI makes attacks faster and cheaper to launch.
The breach-cost signal
IBM said one in four malicious breaches in its study was AI-enabled, a 56% increase from the previous year. It reported an average cost of $6 million for AI-enabled breaches, compared with an overall global average of $4.99 million. Financial-services breaches averaged $6.3 million, according to the company.
PYMNTS and ITPro separately reported the study's spending, deployment, and breach-cost findings. The underlying report was conducted by Ponemon Institute and sponsored and analyzed by IBM, so its figures should be read as IBM's study results rather than independent measurements of every organization or breach.
For ML and security teams, the practical LDS takeaway is to evaluate agent adoption across the entire response chain. Automating detection while vulnerability validation and remediation remain mostly manual can leave the most time-sensitive part of the workflow behind.
Key Points #
- 1IBM reported that 85% of organizations in a May follow-on study planned to increase security spending after learning about advanced frontier AI cyber capabilities, compared with 64% in its broader breach research who planned increases after a breach.
- 2Three quarters said the threats were changing agent deployment across security operations; more than half used agents for detection and containment, but 18% used them for vulnerability management.
- 3IBM reported one in four malicious breaches as AI-enabled and an average $6 million cost, while its sponsorship and survey design limit how broadly the findings should be generalized.
Scoring Rationale #
IBM's survey links frontier-model cyber awareness to planned security spending and exposes a deployment gap between agent-based detection and vulnerability remediation. The findings are useful to ML and security teams, but they describe sponsored survey results and spending intentions rather than a newly demonstrated capability or independently measured market total.
Sources #
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.