AI-Driven Cyberattacks Are Disrupting Security, Warns Expert
Cybersecurity expert Alex Stamos warns AI is rapidly escalating cyberattack risks.
Why it matters: Legal teams depend on digital security to protect sensitive data. As AI automates and amplifies cyber threats, understanding these risks is essential to safeguarding legal operations.
- Alex Stamos, Cognition's chief security officer, warns AI is enabling more frequent, sophisticated cyberattacks.
- Stolen API keys from Fireworks led to malicious Cursor plugins affecting 5,000 companies.
- OpenAI’s GPT-6 Cyber tool identified nearly 1,000 vulnerabilities in the Linux kernel, showing AI’s power to expose security gaps.
- Hybrid AI ransomware recently targeted 47 UK schools, automating and adapting attacks in real time.
Cybersecurity is facing new challenges as artificial intelligence increases the speed and scale of cyberattacks. Alex Stamos, chief security officer at Cognition, describes this shift as a major disruption to established defenses. In a recent interview, Stamos explained how AI enables hackers — including state-backed groups — to find and exploit security flaws more quickly and broadly than before.
AI-based attacks often start with stolen API keys, which are digital credentials that allow software tools to access systems. For example, attackers stole API keys from Fireworks to distribute a malicious plugin via Cursor, infecting about 5,000 companies. This shows how AI automates spreading attacks at scale (AI Daily Journal).
On the defense side, OpenAI developed GPT-6 Cyber, an AI model that scans software code and has identified nearly 1,000 security flaws in the Linux operating system's core software, the kernel. This technology exposes vulnerabilities that both defenders and attackers can exploit, increasing pressure on security teams (Axios).
Cybersecurity researchers also reported a new type of ransomware called hybrid AI ransomware, which uses AI to automate tasks like encrypting files and to adjust tactics during attacks. It recently targeted 47 schools in the UK, demonstrating how AI enables attackers to act faster and more flexibly (IT Pro).
Governments are responding. Japan announced a $63 million emergency cybersecurity fund to strengthen defenses after an increase in AI-crafted attacks on companies and public entities (AP News).
For legal professionals, these developments mean heightened risks to sensitive client data and critical systems. Law firms and legal departments must update security approaches, invest in AI-aware defenses, and train teams to recognize AI-driven attack methods. Staying informed about emerging AI threats is indispensable for protecting legal operations in this evolving environment.
By the numbers:
- 5,000 companies affected — by malware spread via stolen API keys in AI-powered attacks
- Nearly 1,000 security flaws found — in Linux kernel by OpenAI’s GPT-6 Cyber
- 47 schools targeted — in UK by hybrid AI ransomware attack
Yes, but: While AI introduces new threats, it also supports advanced defenses. For example, AI models like GPT-6 Cyber help security teams identify vulnerabilities faster, improving risk management. However, leveraging AI in defense requires expertise and investment to keep pace with attackers.
What's next: Legal organizations should monitor developments in AI-driven cyber threats and adapt security protocols accordingly. Upcoming conferences on AI cybersecurity in 2027 may provide more insights and tools for defense.