Frontier AI Threats Increase Cybersecurity Spending Plans
PYMNTS reported August 14 that 85% of organizations planned to increase security spending in May, up from 64% during the March 2025 to February 2026 period. IBM attributed the increase to greater awareness of frontier AI cyber capabilities, while about 75% of organizations said those threats had prompted them to rethink how they deploy AI agents across security operations.
PYMNTS reported August 14 that 85% of organizations planned to increase security spending in May, compared with 64% during the March 2025 to February 2026 period. PYMNTS, citing an IBM release dated July 29, reported that IBM attributed the increase to organizations becoming more aware of advanced frontier AI cyber capabilities.
The findings also addressed how organizations review their security operating models. According to the IBM release cited by PYMNTS, about 75% of organizations said frontier AI threats had caused them to rethink how they deploy agents across security operations. More than half reported using agents for threat detection and containment, while 18% used them for vulnerability management.
Breach cost and remediation gap
IBM's research, conducted from March 2025 through February 2026, found that AI-enabled attacks represented about 25% of malicious breaches reported in that period, according to PYMNTS. The release also reported a 56% year-over-year increase in the share of breaches enabled by AI.
PYMNTS reported that victims of AI-enabled breaches faced an average cost of $6 million, compared with an overall average breach cost of $4.99 million. For financial-services organizations, the reported average was $6.3 million.
The IBM release linked the operational concern to the interval between vulnerability discovery and remediation. It described deepfake impersonation and other AI-enabled techniques as elements of attacks.
Security spending becomes part of AI deployment
CNBC separately reported that recent AI-related hacking incidents have intensified discussion of cybersecurity as AI labs develop agentic systems. Gene Yu of incident-response firm Blackpanda told CNBC that AI has not changed the volume of vulnerabilities in a system, but can accelerate the rate at which attackers identify them. Paul Meeks of Freedom Capital Markets told CNBC that organizations' cybersecurity outlays could be additive to existing AI infrastructure spending, particularly in finance and healthcare.
For ML and security teams, the IBM findings underscore a recurring industry pattern: using agents for detection without comparable automation in validation, prioritization, and remediation can leave a material response gap. The reported 18% adoption rate for agent-based vulnerability management suggests that deployment maturity may differ substantially across security workflows, even where organizations already use AI for monitoring and containment.
Key Points
- 1IBM survey data cited by PYMNTS shows planned security-spending increases reached 85%, reflecting heightened concern over frontier AI cyber capabilities.
- 2More organizations use agents for detection and containment than vulnerability management, exposing an adoption gap across security-operations workflows.
- 3Industry reporting indicates AI can accelerate vulnerability discovery, making remediation speed and runtime identity controls increasingly relevant to security teams.
Scoring Rationale
The reported survey offers timely evidence that frontier-model cyber capabilities are affecting enterprise security budgets and agent deployment decisions. It is relevant to ML and security practitioners, though it reports spending intentions and survey findings rather than a new technical capability or security vulnerability.
Sources
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.
Try 250 free problems
