AI-driven cyberattacks increase in scale and cost

The cybersecurity landscape is undergoing a fundamental shift as artificial intelligence evolves from a supportive tool into an autonomous agent capable of executing complex attacks. Reports from IBM and Check Point Research indicate that AI-driven attacks are increasing in both frequency and cost. IBM noted a 56% year-over-year rise in AI-enabled breaches, with the average cost of such incidents reaching approximately $6.04 million.

Security researchers highlight a growing ‘speed gap’ between machine-scale offense and human-led defense. While attackers use AI to automate reconnaissance, vulnerability research, and exploit development—sometimes generating functional exploits in under 10 minutes—defenders often rely on manual review and scheduled maintenance. This structural mismatch allows for industrialized exploitation that can bypass traditional signature-based defenses.

Beyond external threats, organizations face internal risks. Research shows that approximately 90% of companies experience high-risk AI tool usage monthly, often involving the accidental input of sensitive data into unauthorized AI platforms. Additionally, the rise of ‘agentic fraud’ leverages AI to conduct highly convincing social engineering and phishing attacks, targeting unverified online channels where human victims are most vulnerable.