5 articles
New role-confusion research explains why prompt injection keeps surviving better prompts. Models do not reliably perceive which text is instruction, tool output, user content, or their own reasoning.
Security researchers showed a €0.02 bank transfer could compromise a banking AI assistant. Here is the exact attack chain - and what every developer building agents needs to do differently.
The ChatGPT for Google Sheets exfiltration report is not just a spreadsheet bug. It is a warning about agentic office tools: permissions need to be action-scoped, logged, revocable, and visible.
Prompt injection stops being an abstract LLM risk once an agent can call tools. The practical defense is data boundaries, structured handoffs, tool guardrails, and approval gates around side effects.
AI coding agents now read repository docs, config, issues, and comments before opening pull requests. That turns CONTRIBUTING.md and AGENTS.md into part of the security boundary.
Showing 4 of 4 articles
New tutorials, open-source projects, and deep dives on coding agents - delivered weekly.
Explore 688 topics
Browse All Topics