Microsoft’s CX strategy is increasingly centred on creating a governed foundation for AI agents that can access enterprise systems and act on behalf of customers and employees.
In its Q2 earnings call, investors saw the fast adoption of Agent 365, hundreds of thousands of Dynamics 365 actions, and deeper investment in data, security and implementation support.
CX leaders must now pay attention to Microsoft’s next CX steps as AI-agent programmes will need credible oversight as much as intelligence, and how it will prove a direct impact on customer and employee outcomes.
Speaking at the earnings call on Wednesday, Satya Nadella, Chairman and CEO for Microsoft, argued that enterprise AI adoption will depend on extending existing governance and security controls to autonomous agents:
“With Agent 365, we offer a control plane that extends companies’ existing governance, identity, security, and management frameworks to agents they build.”
Why Microsoft Is Prioritising Managed Autonomy
Microsoft’s CX strategy follows the in-demand trend of governance, positioning Agent 365 and Dynamics 365 as components of a governed enterprise agent architecture.
From here, autonomous agents can retrieve business context and take action across core enterprise systems while remaining subject to the same permissions and audit controls as human employees.
“We have been reinventing Microsoft Dynamics 365 for an agent-first world,” Nadella announced.
“Agents can now access business context and take action using the same data models, rules, permissions, security guardrails, and audit trails as any application user.”
For enterprise CX , enabling AI agents complete tasks without compromising governance or compliance is the next priority.
To ensure the strategy, Microsoft saw Agent 365 accumulate nearly 40 million registered agents across tens of thousands of companies within two months of launch, and Dynamics now exposes more than 650,000 MCP actions spanning teams including customer service.
The company also pointed to lower inference costs after reporting an 89% reduction in GPU costs in Dynamics 365 through its MAI-Voice-2-Flash model, likely making AI-powered customer service more economical to deploy at scale.
Amy Hood, Chief Financial Officer at Microsoft, observed that as AI agents become embedded in core workflows, enterprises will need clear oversight of both their cost and how they are controlled.
“Everyone is going to need both observability of token spend and the manageability of token spend for all business processes.”
However, despite the focus on governance and operational control, these results do not directly demonstrate better containment rates, faster resolutions, or higher customer satisfaction.
CX leaders must therefore question whether Microsoft’s governance layer will ultimately translate into measurable CX improvements and employee outcomes, or simply remain an enterprise platform advantage.
Governance Alone Will Not Prove CX Value
To become a meaningful CX differentiator, Microsoft’s governance strategy will need to be supported by reliable enterprise data, robust compliance capabilities, and service operations expertise.
To avoid offering governance as a standalone capability, Microsoft has assembled an ecosystem with Foundry, Fabric, Work IQ, Purview, and its Frontier Company services organization to give AI agents access to enterprise knowledge and business systems while maintaining controls.
“The complete app and agent stack. It gives agents access to the IQ layer, the tools they use, along with durable state and memory, secure sandboxes, rubrics and evals, and even their own self-improvement loops,” explained Nadella.
In fact, 100,000 customers now use Foundry, with its revenue more than doubling year over year, as well as tens of thousands of customers grounding agents in enterprise context through Foundry, Fabric, and Work IQ.
“To date, Microsoft Purview has audited over 15 billion Copilot interactions to meet compliance obligations, up nearly 360% year-over-year,” he added.
Despite its governance investments, auditing interactions is a compliance metric rather than a contact center performance measure, creating confusion whether agents are resolving customer issues more effectively, improving satisfaction, or reducing escalations.
As successful AI deployment depends on people as much as technology, the company also plans to embed experts with customers, having already completed more than 330 projects across 164 customers.
Nadella emphasizes:
“We will embed 6,000 industry and engineering experts with customers to co-design, co-innovate, and continuously improve AI systems at scale.”
With enterprise AI implementation requiring ongoing process redesign, CX leaders will need to evaluate whether Microsoft’s governance, data, and services strategy produces measurable improvements in customer service performance and employee productivity.
Until that evidence emerges, Microsoft’s strongest differentiator remains its AI agent infrastructure, instead of proven gains in customer service performance.
