Cyber attacks fueled by artificial intelligence have surged by 89% over the past year, fundamentally reshaping the online threat landscape, according to the latest findings from CrowdStrike. The cybersecurity firm’s 2026 Global Threat Report highlights a rapidly evolving and increasingly aggressive cyber threat environment driven by advancements in AI technology.
The report indicates that AI is not only accelerating the frequency of cyberattacks but also enhancing the speed at which hackers navigate compromised networks. Between 2024 and 2025, AI-assisted cybercriminals increased their operational speed by 65%, with the average breakout time for cyber attacks dropping to just 29 minutes last year. In extreme cases, CrowdStrike documented a record breakout time of 27 seconds, alongside instances of data exfiltration occurring within four minutes.
Hackers are exploiting AI tools to craft more sophisticated attacks while also targeting genuine AI companies. The report revealed that 90 AI firms suffered breaches, resulting in the theft of sensitive personal information and cryptocurrency. Attackers have infiltrated AI development platforms to conduct ransomware assaults and deceive individuals into divulging their credentials by masquerading as legitimate AI services.
Nation-state actors are also increasingly leveraging AI capabilities. The report noted that the Russian-linked group known as FANCY BEAR has utilized large language models to deploy malware, facilitating automated espionage and document theft. Meanwhile, Chinese-affiliated hackers have capitalized on AI advancements, leading to a 38% increase in cyber activity linked to China. The logistics sector emerged as a primary target, with attacks on this industry skyrocketing by 85%. Furthermore, 67% of attacks attributed to Chinese hackers provided immediate system access, with 40% impacting internet edge devices.
In North Korea, AI technology has spurred a staggering 120% rise in cybercrime associated with the regime. The hacking group FAMOUS CHOLLIMA has been particularly active, doubling its campaign frequency and orchestrating the largest financial theft ever recorded, amounting to $1.46 billion in stolen cryptocurrency.
In addition to the rise of AI-driven attacks, CrowdStrike also reported an increase in zero-day and cloud exploits throughout 2025. The data indicated that hackers successfully exploited zero-day vulnerabilities before they were publicly disclosed 42% of the time, while cloud-based cyber attacks saw a 37% uptick.
Adam Meyers, the head of counter adversary operations at CrowdStrike, remarked, “This is an AI arms race. Breakout time is the clearest signal of how intrusion has changed. Adversaries are moving from initial access to lateral movement in minutes. AI is compressing the time between intent and execution while turning enterprise AI systems into targets. Security teams must operate faster than the adversary to win.”
The findings from CrowdStrike underscore an alarming trend in cybersecurity, where the integration of AI into cybercriminal strategies is not merely a technological evolution but a profound shift in how cyber threats are conceived and executed. As organizations and security teams grapple with these evolving tactics, the urgency for enhanced cybersecurity measures becomes increasingly critical to safeguard sensitive data and infrastructure against these sophisticated threats.
See also
Anthropic’s Claims of AI-Driven Cyberattacks Raise Industry Skepticism
Anthropic Reports AI-Driven Cyberattack Linked to Chinese Espionage
Quantum Computing Threatens Current Cryptography, Experts Seek Solutions
Anthropic’s Claude AI exploited in significant cyber-espionage operation
AI Poisoning Attacks Surge 40%: Businesses Face Growing Cybersecurity Risks
