2026: The Year of AI-Assisted Attacks

The article examines a 2024–2026 uptick in AI-enabled cyberattacks—highlighting large data exfiltrations (millions to hundreds of millions of records), a major npm supply-chain compromise (Shai-Hulud), a dramatic rise in malicious packages (hundreds of thousands), and a collapse of traditional exploit and patch timelines—arguing that LLMs/agentic coding have lowered the barrier to sophisticated attacks and recommending structural mitigations such as rebuilding dependencies from verified source (Chainguard Libraries).