Ayelet Kutner, CEO of At-Bay Israel
Artificial intelligence is not replacing traditional cyberattack methods – it is making them more effective. This emerges from the InsurSec report published by cyber insurance company At-Bay for 2026, based on 2025 data.
At-Bay provides integrated cyber insurance and security solutions to approximately 40,000 businesses in the United States (representing combined revenues of 800 billion dollars). The report is based on an analysis of more than 100,000 policy-years of At-Bay cyber claims data.
The report shows that while incidents involving deepfakes and other extreme artificial intelligence scenarios have not yet become a central component in cyber claims, artificial intelligence is already helping attackers execute familiar techniques such as phishing and social engineering with greater precision and efficiency. This is clearly reflected in the data: 52% of cyber incidents began with an email-based attack, highlighting how attackers continue to rely on the most common entry point, now amplified by new technologies that allow these techniques to operate at a broader and more sophisticated scale.
Key findings:
*A 7% increase in the number of cyber incidents compared to the previous year.
*Third-party liability claims surged by 70% in one year and have nearly tripled since 2023.
*The average financial loss from ransomware attacks increased by 16% compared to 2024.
*The technology sector leads in ransomware damages, with losses 72% higher than the overall average.
*The use of remote access as an entry point for ransomware attacks has nearly doubled within two years.
