Global Cyberattacks Average 1,900 Per Week as GenAI Data Risks Rise: Report

Organisations worldwide faced an average of 1,900 cyberattacks per week in September 2025, according to Check Point Research’s Global Threat Intelligence Report.

While that’s a 4% dip from August, it marks a 1% increase year-over-year, underscoring that global cyber threats remain at historically high levels. The report highlights an alarming trend linked to Generative AI (GenAI) use across enterprises.

Check Point found that 1 in every 54 GenAI prompts posed a high risk of sensitive data exposure, affecting 91% of organisations using GenAI tools regularly. Another 15% of prompts contained potentially sensitive information such as customer data, source code, or internal communications — revealing growing gaps in AI governance and data protection.

“Ransomware remains the most destructive force, while the emergence of GenAI-related data leakage adds a new dimension of risk for organisations. Cybercriminals will likely seek to exploit every innovation faster than users can adapt,” Omer Dembinsky, Data Research Manager at Check Point Research, said.

The education sector remained the most targeted globally, with an average of 4,175 weekly attacks per organisation, reflecting its expanding digital footprint and limited cybersecurity funding. Telecommunications and government institutions also saw significant attack volumes, averaging 2,703 and 2,512 weekly incidents, respectively.

Regionally, Africa experienced the highest attack frequency, followed by Latin America and Asia-Pacific, while North America saw a 17% surge driven largely by ransomware.

Ransomware remained the most disruptive threat, with 562 publicly reported incidents — up 46% year-over-year. North America accounted for more than half of these cases, led by the U.S. (52%).

"The only sustainable defense is a prevention-first strategy powered by real-time AI, ensuring protection across the network, cloud, endpoints, and identities," Dembinsky added.