Big Tech's pivot toward agentic AI represents a fundamental shift in where value extraction happens within customer relationships. Meta's launch of Business Agent exemplifies this move: rather than stopping at advertising delivery, the company now captures the transactional moment itself—answering questions, qualifying leads, processing bookings and payments directly within WhatsApp and Instagram. This mirrors parallel plays from Microsoft (embedding agents into Dynamics 365), Amazon Web Services (leveraging cloud infrastructure for enterprise automation), Google (integrating agentic capabilities into search and workspace tools), and OpenAI (pushing custom multi-agent frameworks). The market opportunity is substantial: agentic AI is projected to grow from $10.9 billion in 2026 to $182.9 billion by 2033. What distinguishes this battleground from previous platform wars is that these companies are converging on the same prize—not just customer attention, but the mediation layer between businesses and their customers.
For CX teams, the implications are dual-edged and demand strategic clarity. The immediate benefits are tangible: AI agents handle routine inquiries across languages, free staff for complex issues, and deliver capabilities previously available only to large enterprises. Yet this convenience masks a structural power shift. As interactions become increasingly mediated by platform-owned AI, businesses risk dependency on systems they cannot control, whilst platforms accumulate unprecedented insight into customer behaviour, hesitation points and purchase drivers. For teams already running Agentforce or similar enterprise solutions, the question becomes whether your vendor's agent architecture insulates you from platform lock-in or accelerates it. The real risk emerges over time: as more customer interactions flow through these agents, the platforms mediating them gain disproportionate influence over what information customers see, which products get recommended, and how journeys progress from enquiry to purchase.
The governance challenge is equally pressing. With 96% of organisations reporting that agentic AI deployments met or exceeded ROI expectations, adoption will accelerate—but without clarity on who controls the agent's decision-making logic, how customer data flows, and what commercial value platforms extract from interactions. CX leaders must distinguish between agents that augment their control over customer relationships and those that gradually transfer that control upstream to platform providers. The battleground is not really about automation; it is about who owns the customer relationship itself.
More than chatbots: why business AI agents are Big Tech’s next product battleground The Conversation