Latest release of agent-building platform bolsters orchestration in complex workflows, simplifies integration by tapping Model Context Protocol (MCP).
governance
Microsoft’s first ‘autopilot’ ushers in an era where IT and business leaders must decide what an always-on agent can touch, act on, and bring to users’ attention.
Agent Optimizer functions ensure Foundry Hosted Agents adhere to user-defined criteria for the best instructions, skills, models, and tool descriptions.
Autopilots increase the level of AI autonomy and automation. Build event also places major focus on agentic AI development — and controls — at the desktop.
Microsoft updates Finance Agent to provide a command center-type interface, as well as support for natural language prompts to gain and act on financial insights.
Support for Compliance API gives Purview visibility into conversations and events, bringing both data types in Claude under broader enterprise controls.
Features for controlling the actions of agents — ensuring they don’t misbehave — bring the enterprise-level controls required by IT/security pros and business leaders alike.
At Paragon Films, MCP is enabling AI agents to do everything from managing approval workflows to pinpointing invoice discrepancies to making employee knowledge ‘immortal.’
Skills codify workflow details that are too often managed as ‘institutional knowledge,’ making those instructions accessible and reusable for agentic AI automation.
As initial release of ‘control plane’ becomes widely available, Microsoft is expanding agent registry, Shadow AI protections, and threat detection capabilities.
The amended partnership shows both companies are evolving; it’s not a falling out. The question for organizations is whether they are ready to make similar strategic shifts to focus on governance and outcomes.
Deloitte partner says most organizations are still in the transition phase of tapping AI in their ERP environments, but the technology can improve data consistency and improve financial compliance.
Updates include learned context and skills that get applied to future sessions as well as automatically created content that adheres to corporate standards.
Toolkit acts as a gatekeeper, preventing unauthorized access by AI agents and addressing top risks including goal hijacking and memory poisoning.
Latest MCP product from Microsoft works with cloud or on-premises SQL databases for efficient, secure connections to corporate resources without requiring language or framework expertise.
AI initiatives will fall for those companies that focus on the model and interface without building a solid underlying foundation to drive tangible business value.
Round of enhancements also includes simplified prompting during agent-building processes, new content moderation controls to govern sensitive material.
Initial release supports actions across agents, multiple AI models, and interoperability through the two most widely embraced interoperability standards.
New metrics provide visibility into usage patterns, while expanded controls improve governance over content sources in Purview platform.
The company and ISV partners deliver agents that inject new functions into Copilot, Defender, and other Microsoft platforms, driving significant growth in the security ecosystem.