Urgent.News

What's breaking now, across thousands of outlets.

AI

Building a Switch courtroom where AI can negotiate but can't consent

A creative lead wants someone's portrait in a promo. That person wants to contribute, but doesn't want their face used. An AI can suggest a compromise. Who gets to say yes? I'm Harshil, a B.Tech AI/ML student in Delhi. My teammates Daksh, Kartik and I built Delegates Courtroom for Flip the Switch (Track 2: Best Personal or Creative Workflow). Two AI advocates negotiate a storyboard, and the…

Two AI advocates - named cvc-lead-advocate and cvc-creator-advocate - negotiate a storyboard for a promotional video. They adhere to the rule that human sign-off is required, while the AI handles negotiations. The Switch platform provides a shared workspace for agents to collaborate on the project, with a recorded rehearsal showcasing the negotiation process.

The original rehearsal involved synthetic human approvals, but the final output demonstrates the mechanism's functionality, without any actual human agreement. The project was built using a pinned Docker stack, Mattermost chat app, and a local polling adapter.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

My First Switch Hackathon: Setting Up an AI-Powered Workspace

My First Switch Hackathon: Setting Up an AI-Powered Workspace Introduction For the Flip the Switch Hackathon, I decided to explore how an AI-powered workspace could help organise and manage creative…

  • Installed Switch on Windows machine, encountered tools and concepts
  • Integrated PostgreSQL database, connected to OpenCode AI coding agent
  • Faced authentication failure, learned about version discrepancies and troubleshooting

Designing an AI incident agent for 3 a.m.: making memory visible

Picture the person who uses an incident tool. It's the middle of the night. An alert just woke them up. Customers are affected, a manager is asking for updates, and they are trying to read logs with…

  • DejaVu AI agent provides answers and proof for off-hours incident handling
  • Live memory map visually displays past incidents grouped by type and resolution time
  • Comparison feature shows memory's impact on confidence with confidence level proof

Can an MCP Server Actually Help Prevent AI Agents From Leaking Secrets?

I've been experimenting with MCP servers and AI coding agents while building an open-source project called StagedSec. One thing I've been thinking about is where security checks should happen when an…

  • StagedSec project aims to prevent AI agents from leaking sensitive information.
  • MCP server exposes scanning capabilities to AI agents for secret detection.
  • Challenges include handling split secrets, obfuscated content, and security enforcement.

More from Saturday 7 November →