IT security company Check Point Software has released its first edition of the AI Security Report. The report provides an in-depth analysis of how cybercriminals leverage AI to carry out attacks, along with actionable insights on how organizations can defend against these evolving threats.
As AI continues to reshape industries and become more deeply integrated into business operations, the line between truth and falsehood online is blurring. The report finds that a majority (51%) of companies already use AI in daily operations. Criminals exploit generative AI and large language models to undermine digital trust. What you see, hear, or read online can no longer be assumed true. AI-driven impersonations can bypass even advanced verification systems, putting individuals and organizations at greater risk.
“Cybercriminals have adapted quickly to using AI, and that is already changing the threat landscape,” says Mats Ekdahl, security expert at Check Point Software. Security teams must match this pace by integrating AI into their own defenses. The report highlights not only the risks but also recommended practices for securing AI environments.
Key findings from the report:
- AI-enhanced impersonation and manipulation
Cybercriminals use AI to create realistic digital identities, generate live phishing messages, and blur the distinction between authentic and fabricated content. - AI-generated malware
Attackers employ AI to design and optimize malware, automate DDoS campaigns, and refine stolen credentials for greater effectiveness. - AI use and data-leak risk
Analysis of Check Point’s GenAI Protect data indicates that about 1 in 80 prompts carries a high risk of leaking sensitive information. Overall, roughly 7.5% of prompts (about 1 in 13) contain sensitive data, creating serious security and data-integrity challenges as AI adoption grows. - Detect threats with AI
Use AI defensively to identify AI-generated threats such as phishing and deepfakes, improving detection speed and accuracy. - Stronger identity verification
Move beyond traditional verification methods and adopt multi-layered identity checks that account for AI-driven impersonation across text, voice, and video.
Download the full report here.
Read more on Check Point’s blog: https://blog.checkpoint.com/research/ai-security-report-2025-understanding-threats-and-building-smarter-defenses/