GAO says AI can magnify cyberattacks and strengthen cyber defense
2 min readThe Government Accountability Office said in a report published Friday 20 November 2026 that AI technologies have given threat actors the ability to magnify the reach and impact of traditional cyberattacks. The agency said AI makes it possible to create high-quality phishing and deepfake content, improve the precision and accuracy of attacks, and accelerate the speed and scale of malicious cyber activity.
According to the report, agentic AI was among the emerging trends in AI-enabled cyberattacks. It can autonomously carry out multiple stages of an attack with minimal human input.
In June 2026, a threat actor used AI to build a testing framework for refining malware techniques intended to evade detection. GAO said investigators found that multiple AI agents tested malicious code against different security products. The agents searched security research for new attack techniques, mapped them to known cybersecurity frameworks, conducted attacks in test environments and reported the results.
In Feb. 2026, a low-skill level threat actor used multiple commercial generative AI services in a widespread attack. The actor compromised several devices by using stolen credentials from past incidents on devices with weak passwords and no multi-factor authentication enabled, according to the report.
The report said cyber defenders use AI, including machine learning and generative AI applications, to safeguard information systems and improve security tools, teams and operations. Defenders use AI to analyze network traffic, detect vulnerabilities in source code and parse large amounts of network information quickly. Integrating generative AI into tools can support dynamic scanning that factors in new known vulnerabilities and environment activities.
GAO identified four potential policy goals: address gaps in cybersecurity guidance and practices, address AI-related research gaps, govern AI tools throughout their cybersecurity lifecycle and enhance the cybersecurity workforce. It also identified 16 policy options for policymakers, including requiring human authorization for consequential AI-initiated cybersecurity actions, validating AI cybersecurity tool output and strengthening AI expertise in the cybersecurity workforce. The agency said experts also identified data privacy, quality, technical, governance, trust, cost, skills, integration, regulatory and data-sharing challenges.



