Big Tech companies including Microsoft, Google, Nvidia and Amazon are prioritising agentic artificial intelligence (AI) in 2025 and driving innovation in the technology as a result, a new report outlines.

GlobalData’s Agentic AI report notes that various Big Tech companies have been rolling out developer-oriented platforms and frameworks in recognition of the developer community as a co-architect of agentic AI.

“These platforms enable developers to build customised AI agents across various domains, from customer service and logistics to cybersecurity and creative design,” the report details. “It also means that AI software vendors do not have to supply such customisation in-house.

“This approach mirrors the early development of the internet in the 1990s, during which time developers were crucial for creating new applications and functionality, rapidly advancing future internet-dependent innovations. Similarly, developers have been central to the creation of apps for iOS and Android mobile operating systems.”

Agentic AI is a form of the technology that is able to act autonomously in making decisions and taking actions. Typically able to communicate and collaborate, AI agents are being developed for various consumer, enterprise, scientific and industrial purposes, with early adopters in the healthcare, financial services, manufacturing, scientific research, energy and utilities sectors.

Agentic AI use cases

Examples include Affineon, which automatically handles lab results and prescription renewals to save time and improve patient care, and HCLTech, which detects defects and anomalies in different industries, with real-time dashboards and virtual assistance to reduce costs and improve production.

Among the Big Tech moves advancing agentic AI, the report points to Microsoft’s Magnetic-One multi-agent architecture, AutoGen open-source framework for building AI agents, Semantic Kernel software development kit and AI Workbench development environment.

Of Google’s efforts, it says: “Project Astra, Vertex AI Agent Builder and Gemini 2.0 Flash expand the Google Cloud AI agent ecosystem. Google Agentspace provides AI agents with multi-model search for enterprises, first-party ingest, blended retrieval-augmented generation and third-party connectors. Google has announced Memory Bank for Vertex AI Agent Engine, a persistent-memory feature that lets AI agents remember long-term context and user preferences across sessions.”

Agentic AI startups are also driving innovation through diverse use cases. Ada specialises in agentic AI for customer service and resolving customer queries; Ema provides agents that augment the employee experience; and AIQ tackles issues like drilling performance, reservoir management and quality control in the oil and gas sector.