IBM Report: AI-Driven Cyberattacks Are Escalating
A new report from IBM's X-Force finds that cybercriminals are increasingly using AI tools to accelerate attacks by identifying security weaknesses more rapidly. The 2026 Threat Intelligence Index highlights that basic security gaps are being exploited at a dramatically higher rate due to these AI-driven methods.
- For the fifth consecutive year, manufacturing was the most targeted industry, accounting for 27.7% of all observed incidents, with data theft being the most common objective. - Vulnerability exploitation has become the leading initial cause of attacks, responsible for 40% of the incidents IBM's X-Force responded to in 2025. This was followed by a 44% increase in attacks that began by exploiting public-facing applications. - Generative AI has drastically reduced the time needed to create convincing social engineering attacks; an effective phishing email that would take a human team 16 hours to research and write can now be generated by AI in just five minutes. - Phishing remains a dominant threat, with some reports attributing a 1,265% surge in these attacks to the use of generative AI tools that eliminate tell-tale signs like grammatical errors. - The number of active ransomware and extortion groups surged by 49% year-over-year, indicating a fragmentation of the ecosystem, while publicly disclosed victim counts rose by approximately 12%. - Large-scale compromises of supply chains and third-party services have nearly quadrupled since 2020 as attackers increasingly target software development environments and SaaS integrations. - Credential theft is a primary goal for attackers, with over 300,000 ChatGPT credentials found exposed on the dark web in 2025. This is part of a broader trend where infostealing malware is on the rise as a method to scale identity-based attacks. - Defenders are also using AI to counter these threats by analyzing vast amounts of data to detect anomalies, predict attacks, and automate responses faster than humanly possible.
