This week saw major strides in AI agent collaboration across industries. In New York, security experts gathered at the AI Agent Security Summit to discuss defense strategies against malicious AI agents working together. Meanwhile, Microsoft unveiled Sales Agent, a new AI tool that teams up with CRM systems to prioritize customer leads automatically.

Researchers in Paris demonstrated how medical AI agents can predict heart attacks 3 hours early by analyzing patient data together. Schools aren't left out - Georgia Tech students showed AI agents helping teachers grade papers faster through team-based checking systems.

A big debate emerged about Klarna's plan to replace hundreds of workers with AI agent teams, raising questions about job impacts. On the bright side, Stanford University revealed Collaborative Gym, a training system helping humans and AI agents work better as partners.

Tech companies like Cisco and Jotform launched new teamwork-focused AI agents for customer service and form processing. Chinese researchers shared open-source tools making AI agents better at group problem-solving in different languages. As these smart helpers learn to cooperate, experts warn we need strong safety rules to keep them helpful.