Eighty percent of ransomware attacks now use artificial intelligence, but AI cybersecurity is catching up, according to a new report.
Summary
- A new report from OutreachX shows that artificial intelligence now fuels 80% of ransomware attacks, supercharging cybercrime with tools like large language models and deepfakes.
- Yet the same report highlights how defenders are fighting back with AI of their own, cutting costs, compressing response times, and turning the battle into a high-stakes arms race between malicious code and machine-powered defense.
Cybersecurity in 2025 is no longer just human versus human: it’s AI versus AI. According to a report by OutreachX, cyber criminals are increasingly using artificial intelligence tools to streamline and scale their activities. On the other hand, cybersecurity experts are increasingly utilizing AI to address this emerging threat.
The report noted that AI is contributing to an increase in ransomware attacks. Notably, in 2021, there were an estimated 7,850 attacks daily. However, by 2031, this figure is projected to grow to 43,200 daily incidents. AI is a significant contributing factor to this explosion, helping to scale and automate the once manual and skill-intensive attacks.
What is more, AI now plays a role in 80% of ransomware attacks, through technologies such as LLMs and deepfakes. Fake videos are alarming, as potential victims could identify AI-generated voices only 60% of the time.
You might also like: July crypto hacks top $140M among 17 major exploits: PeckShield
AI increasingly plays a role in cybersecurity
Still, to tackle this issue, companies increasingly leverage AI tools. This includes machine learning to detect suspicious activity, automating phishing triage, and accelerating threat responses. According to IBM data, AI tools in cybersecurity save $1.9 million per breach. By 2027, this technology is projected to save enterprises $10.4 billion.
“The biggest unlock isn’t a flashier dashboard, it’s compressing dwell time. When AI closes the loop from detection to containment in minutes, board-level metrics move: fewer payouts, faster recovery, and less revenue at risk. For mid-market teams, automating phishing triage and identity anomalies delivers the fastest ROI by cutting alert fatigue and freeing analysts to hunt, not babysit queues,” said Anirudh Agarwal, CEO, OutreachX.
Read more: $2.4B lost in 2025 H1 crypto hacks — exchanges and DeFi hit hardest: report
