The race to build AI agents heated up this week as companies worldwide revealed new tools and strategies. NTT DATA (based in Japan and the U.K.) made waves with its Smart AI Agent Ecosystem, designed to offer tailored automation for industries like healthcare and finance. Their patented system can turn old robotic process automation (RPA) bots into intelligent agents, saving costs for businesses. The firm also plans an OpenAI Center of Excellence to accelerate development, signaling strong commitment to the space.

In the U.S., Microsoft outlined a vision where AI agents from different companies work together seamlessly. For example, a shipping company’s agent could automatically share data with a supplier’s agent to avoid delays. This approach could help startups specializing in niche tasks integrate into larger systems.

However, not all news was positive. Venture capitalists at a16z admitted that the term "AI agent" remains poorly defined, causing confusion in the startup ecosystem. This ambiguity makes it harder for new companies to explain their value. Established players like IBM added to the pressure by rolling out their own agent-focused tools, raising the bar for innovation.

On the product side, UiPath (a U.S.-Romanian automation firm) launched AI agents within its Maestro platform, letting businesses manage both human and AI workflows in one place. This move highlights how startups might partner with larger platforms to reach customers.

Globally, the push for agentic AI is creating opportunities but also challenges. Startups must navigate unclear definitions, compete with tech giants, and prove their agents can deliver real-world results. Partnerships—like those hinted at by Microsoft—could be crucial for survival in this fast-changing field.