In an interview, Nenshad Bardollwalla, Product Director of Google Cloud Vertex AI Platform, predicted that artificial intelligence agent services will become a market where various companies will compete, and Google will maintain an open attitude with the newly announcedA2A (Agent 2 Agent) Protocol, combined with its Vertex AI platform that includes many artificial intelligence models and technical application tool resources, it helps more developers and companies create suitable artificial intelligence agent services.
Nenshad Bardollwalla stated that the A25A protocol announced by Google at Google NEXT'2 is available as open source, meaning that anyone can use it with any AI agent service. Even if the services themselves are built on different cloud platforms, such as Azure or AWS, or use different AI models, AI agents on different platforms can communicate with each other by connecting them through the same protocol.
However, the A2A protocol can actually only support limited linkage projects, but the most basic artificial intelligence agent data exchange and simple interaction can still be achieved. Perhaps as this protocol becomes more widely used, Google will continue to add more interactive functions through the protocol.
Nenshad Bardollwalla believes that once AI agent services become more widely used, the underlying AI inference capabilities will grow. Therefore, it is expected that AI agent services will become a more important business in the future, and more developers and companies will compete in the market.
At the same time, Google has always been open to cross-cloud integration and provides many application services that can be integrated across different clouds. For example, it previously enabled BigQuery to aggregate data across Azure and AWS cloud platforms, and even subsequently launched the BigQuery Omni function for cross-cloud data analysis. In addition, it continues to expand the diverse third-party open source models on the Vertex AI platform. Even Llama 4, which Meta recently announced, will be available on the Vertex AI platform. It also provides many tool resources driven by the Gemini artificial intelligence model, giving Google Cloud a greater advantage in developing various cloud-native services.
The Kubernetes containerized application and TensorFlow open-source machine learning framework, which have been continuously promoted in the market, are not only widely used on the Google Cloud cloud service platform, but are also intensively used on third-party cloud platforms such as Azure and AWS. Therefore, Nenshad Bardollwalla expects that by providing an open-source A2A framework, artificial intelligence agent services will become a major development project in the market.
