Autopilots increase the level of AI autonomy and automation. Build event also places major focus on agentic AI development — and controls — at the desktop.
security
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.
As initial release of ‘control plane’ becomes widely available, Microsoft is expanding agent registry, Shadow AI protections, and threat detection capabilities.
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.
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.
WorkIQ-powered Copilot Cowork orchestrates complex tasks to the point that agents could take over one-fifth of any team’s work in the near term, while hastening the end of app-driven workflows.
This session outlines emerging best practices for securely governing unstructured data in AI systems, showing how identity, permissions, metadata, and retention work together to enable compliant, controlled retrieval.
This session focuses on how to operationalize Copilot Studio agents at scale, covering the controls, governance, lifecycle management, and telemetry needed to run secure, production‑ready, multi‑agent programs.
This session explores how AI connects to ERP and enterprise systems, focusing on security, safer integrations, and minimizing data exposure across connected apps and copilots.
This session shows how the Model Context Protocol streamlines AI adoption while strengthening data‑driven decisions, security, and compliance.
Dive into the Copilot Control System in this session covering the latest Microsoft 365 Admin Center updates and other Copilot features.
This session explores Work IQ, focusing on its architecture, security and privacy design, and built‑in governance capabilities.
Anthropic Claude-powered feature headlines Copilot Wave 3 announcement. Other enhancements include app-native Copilots and the Microsoft 365 E7 ‘Frontier Suite.’
As enterprises require AI models to deliver consistent, secure performances on a regular basis, organizations need an AI-ready security foundation that protects their data, meets regulatory demands, and brings order.
Recent events in the US and Canada highlight the privacy-public safety dichotomy, as well as questions on the responsibility of AI firms in sharing signals of future harm.
New products deliver visibility, governance, and data protection controls over agents built internally as well as those from Microsoft and third parties.
Security Dashboard for AI is built to rein in AI sprawl with Defender, Entra, and Purview integration as well as inventory functions that span Microsoft and widely used third-party AI software.