Global Digital Times

Rules fail at the prompt, succeed at the boundary

Global Digital Times • January 28, 2026 • 5 min read
Featured image for Rules fail at the prompt, succeed at the boundary

Rules fail at the prompt, succeed at the boundary

From the Gemini Calendar prompt-injection attack of 2026 to the September 2025 state-sponsored hack using Anthropic’s Claude code as an automated intrusion engine, the coercion of human-in-the-loop agentic actions and fully autonomous agentic workflows are the new attack vector for hackers. In the Anthropic case, roughly 30 organizations across tech, finance, manufacturing, and government were…

Source: MIT Technology Review