The Linux Foundation’s AGNTCY project emerged as a key initiative this week, aiming to create a universal infrastructure for AI agents. Launched by Cisco in March 2025 and now supported by Dell, Google Cloud, Oracle, and Red Hat, AGNTCY provides tools for agent discovery, secure messaging, and observability. This means AI systems from different vendors can now find each other, communicate safely, and share performance data—critical for building complex multi-agent systems. For example, AGNTCY integrates with Anthropic’s Model Context Protocol (MCP) and the Agent2Agent (A2A) project, enabling agents to collaborate across platforms while maintaining transparency.

Stanford Medicine researchers introduced virtual scientists, AI agents designed to mimic human researchers. These agents work in teams, using language models to discuss hypotheses, analyze data, and propose experiments. In a test case, they rapidly developed a novel approach for a SARS-CoV-2 vaccine, outperforming some human-led efforts. The system’s strength lies in its ability to combine specialized knowledge—like virology and immunology—through agent collaboration, mirroring how interdisciplinary teams solve real-world problems.

At the WAIC 2025 conference, GPTBots unveiled a platform to transform standalone AI tools into collaborative teams. Enterprises can now create custom agent networks where tools like chatbots, data analyzers, and workflow managers share tasks and insights. This approach addresses the limitations of single-purpose AI systems by enabling them to work together, much like human teams. For instance, a marketing team could use GPTBots to connect a language model for content creation with a data analysis tool for trend forecasting, automating complex workflows.

These developments reflect a broader trend toward agentic AI, where systems act autonomously but also collaborate. The AGNTCY project’s focus on interoperability ensures agents from different ecosystems can work together, while Stanford’s virtual scientists and GPTBots’ platform demonstrate practical applications in research and business. Together, they signal a future where AI agents function as cohesive teams, driving innovation across industries.