Hackers ‘Learn in Real Time’ As Agentic AI Drives 36% Uptick in Found Vulnerabilities

By | August 18, 2026

The use of agentic artificial intelligence is benefiting research in cyber vulnerabilities faster than it is used to exploit them—so far, according to new research.

Beazley Security’s “Q2 2026 Quarterly Threat Report” found agentic AI used in research drove a 36% quarter-over-quarter increase in newly disclosed vulnerabilities.

However, vulnerabilities confirmed as actively exploited and added to the Cybersecurity and Infrastructure Security Agency’s (CISA) Known Exploited Vulnerabilities catalog rose only 10% during the same time.

“The headline this quarter is that AI made the security industry’s job noisier without making the attacker’s job fundamentally different,” said Alton Kizziah, CEO of Beazley Security, in a statement. “But AI assisted attacks are gaining in both frequency and effectiveness, and we seem to be watching the attackers learn in real time.”

Beazley Security said attackers are experimenting with AI but thus far are relying on tried-and-true methods, such as compromised credentials, to hack organizations. The use of compromised credentials accounted for 67% of ransomware intrusions investigated by Beazley Security during Q2. Business email compromise also remained a common incident type, said the cybersecurity services firm.

“The front door has not changed. The noise around it has,” Beazley Security said with further advice for organizations to remain vigilant with cybersecurity hygiene and consider assessments to determine how AI is used within a particular enterprise.

Both total new vulnerabilities and high-risk vulnerabilities surged in Q2, with over 20,700 new vulnerabilities that were publicly disclosed. Of those, nearly 5,600 were considered high risk and could be exploited remotely with potential to cause harm if unpatched, Beazley Security reported. Twenty-one of the high-risk vulnerabilities prompted Beazley Security to issue advisories to customers in Q2. That is an “alarming” 40% increase compared with the previous quarter, the firm added.

Topics InsurTech Data Driven Artificial Intelligence Cyber Agencies

Was this article valuable?

Here are more articles you may enjoy.