AI-driven cyberattacks surge 56% and now represent quarter of all data breaches

An IBM and Ponemon Institute research study found that artificial intelligence-driven attacks represented one-quarter of all data breach incidents, marking a 56% increase from the previous period. Each AI-powered attack causes average losses of $6 million, 20% higher than the average cost of traditional data breaches. Sixty-two percent of AI-driven attacks target critical infrastructure, concentrating on financial and energy sectors. This pattern elevates risks of cascading impacts across the broader economy, supply chains, and essential services. Only one-quarter of organizations deploy AI and automation to mitigate breach costs by approximately $2 million per incident. The study identifies significant defensive gaps: merely 18% of enterprises have deployed AI agents for vulnerability remediation, though over half already use AI for threat detection and containment. More than one-fifth of organizations experienced intrusions specifically targeting AI models and systems, often stemming from inadequate cloud environment configurations and weak perimeter defenses.