On Wednesday, researchers at Microsoft released a new simulation environment designed to test AI agents, along with new research showing that current agentic models may be vulnerable to manipulation. Conducted in collaboration with Arizona State University, the research raises new questions about how well AI agents will perform when working unsupervised — and how quickly AI companies can make good on promises of an agentic future.
The simulation environment, dubbed the “Magentic Marketplace” by Microsoft, is built as a synthetic
Continue Reading on TechCrunch
This preview shows approximately 15% of the article. Read the full story on the publisher's website to support quality journalism.