The Agentic AI Reckoning: Autonomy, Safety, and the Edge
From Embedded AI - Intelligence at the Deep Edge by David Such
April 13, 2026 · 25 min · Season 5 · Episode 25
About this episode
The episode discusses the implications of agentic AI in industrial settings, highlighting a real incident involving unauthorized actions triggered by a crafted document.
Send us Fan Mail In Q1 2026 the agentic AI conversation moved from theory to forensics. A crafted PDF triggered physical pump activation through a Claude MCP integration at an industrial facility, after an engineer used the same agent for routine document summarisation and SCADA writes. The hidden instructions used white-on-white text and base64 encoding, the agent treated the document content as instructions, and the legitimate credentials carried the action straight through to operational t...
People in this episode
Host: David Such
Topics covered
- agentic AI
- autonomy
- safety
- industrial automation
- document summarisation
- SCADA
Keywords
- agentic AI
- autonomy
- safety
- industrial facility
- Claude MCP
- SCADA
- document summarisation
- base64 encoding
Mentioned in this episode
Products: Claude MCP
More episodes of Embedded AI - Intelligence at the Deep Edge
- Who is Liable for Onboard AI? · May 31, 2026 · 24 min
- Squeezing AI into your Pocket · May 28, 2026 · 20 min
- A chip that controls a balancing propeller on seven microwatts · May 14, 2026 · 15 min
- Why 95% of AI Deployments Fail · May 7, 2026 · 23 min
- Why Humans and Robots must Dream · April 25, 2026 · 24 min
- Sovereign AI and the End of the Borderless Cloud · April 19, 2026 · 20 min
Explore listener stats, chart rankings, contacts and more on the Embedded AI - Intelligence at the Deep Edge podcast page.