New agents, playbooks, AI-readiness initiatives and vulnerability shielding help security and IT teams go from finding exposure to verified risk reduction
Security teams face more vulnerabilities, misconfigurations and attack paths than they can triage by hand. With the rise of AI tools, attackers can exploit a vulnerability in as little as 35 minutes after it was published (source), making the window to address new CVEs and apply patches smaller than ever. The existing approach to exposure management can’t keep up with addressing risk.
To address this, we are bringing delivering a new set of agentic exposure management capabilities in Microsoft Perception, helping organizations reduce the risks that matter most today, while strengthening their security foundation.
There are three ways customers can benefit through these new innovations: agents, playbooks and AI-readiness initiatives.
- Green agents specialize in specific security tasks, such as identifying and prioritizing exposures, determining which actions will reduce the most risk, and developing actionable risk-reduction plans.
- Playbooks orchestrate agents to accomplish specific security outcomes through repeatable workflows,
- AI readiness initiatives provide structured, ongoing programs to strengthen security foundations and track risk reduction over time. They bring together recommended controls, measurable outcomes, and coordinated human and agent actions to address priorities such as end-to-end vulnerability management and modernizing authentication.
Automate posture improvement with Green agents
Agents built to help accelerate exposure management:
- Posture Prioritization Agent, now Generally Available, helps security teams decide which posture risks to address first ,and why, by bringing together available Microsoft Defender signals and context, including exposure, exploitability, reachability, and asset criticality. It explains why each risk is prioritized, connects it to affected assets, and provides actionable remediation guidance, while leaving execution to separate governed workflows.
- Risk Reduction Agent, now in Preview, takes a prioritized exposure and creates an actionable risk-reduction plan. It coordinates the appropriate remediation or mitigation path through other agents and capabilities, preserves human approval for critical actions, and tracks progress toward closure.
- Additional agent capabilities:
- Vulnerability remediation: Connects vulnerability intelligence with Microsoft Intune remediation. It turns prioritized endpoint risks into coordinated patching and controlled deployment plans that endpoint administrators can review and execute.
- Vulnerability shielding: Vulnerability shielding provides targeted, temporary protection for vulnerable systems while permanent remediation is underway, with operational-impact analysis and human approval built into the process. Shielding and patching work together as complementary interventions, helping organizations reduce risk immediately while moving toward a durable fix. This is now Public Preview for server-side RCE vulnerabilities, a scenario we see for 37% of all exploits. We plan to continue further expand the capability. Read here to learn more about shielding during an active attack.
- Additional agent capabilities:
Deliver security outcomes through playbooks
Playbooks help coordinate agents through dynamic workflows that orchestrate specialized agents to execute security tasks, with human oversight. Customers run playbooks to initiate work.
The Plan for posture remediation playbook, now generally available, uses the Posture Prioritization Agent to turn posture findings into an ordered remediation plan.
Signals: The agent draws on signals such as cloud insights from Microsoft Defender for Cloud with Defender Cloud Security Posture Management.
Scoring: Each exposure is scored on likelihood, potential impact, asset criticality, exploitability, reachability and the number of affected resources. The agent also looks for single fixes that reduce risk across several assets or attack paths.
Result: A ranked list of the most important issues. For each one, the plan explains why it matters, names the critical assets involved and recommends the specific control or configuration change needed to close the gap.
- The Reduce exposure risk playbook puts that analysis into action. It uses the Risk Reduction Agent to evaluate the options for reducing risk and recommend the best response.
- When a permanent fix is ready the agent can coordinate patching through Microsoft Intune while IT keeps control of deployment.
- When patching isn't immediately possible or could disrupt the business the playbook can help shield vulnerabilities with targeted protection while teams work toward the permanent fix.
Set secure foundations with AI-Readiness Initiatives
Coming to preview in January, AI Readiness Initiatives help organizations prepare for AI adoption and transformation by tracking progress on critical security work overtime. Building on guidance already available in Secure Now, each of these initiatives help give agents and security practitioners a shared, structured plan for closing gaps in a specific area. Progress is measured using live tenant data, providing evidence that gaps have been closed and exposure has been reduced.
Each initiative begins with a plain statement of what must be true before an organization can safely scale AI—for example, requiring everyone to use modern authentication. These requirements reflect the attack patterns behind real-world breaches, the ways AI is accelerating those patterns, lessons from Microsoft's Secure Future Initiative, Zero Trust principles, and customer input.
How the initiatives work
AI Readiness Initiatives translate each requirement into recommended controls, readiness tasks, and measurable outcomes. As teams complete the work and strengthen their posture, they can track how risk declines. Every initiative follows four steps aligned with the continuous threat exposure management (CTEM) framework:
- Scope & discover: Review readiness by focus area and choose what
to harden next. - Prioritize & validate: Understand what needs to change in your estate, and where.
- Mobilize & remediate: Agents, automation or guided steps help you make the changes.
- Verify & report: Metrics show your posture improving and the risk that remains.
Highlights include:
- Patching and vulnerability management: Assess, shield and coordinate patching for priority vulnerabilities. This initiative can be used to address large volumes of patches that come out in scenarios like a monthly Windows update.
- Modernize authentication: Strengthen administrator identity security by guiding the rollout of phishing-resistant passkeys.
- Agent discovery and governance: Help organizations strengthen their security foundations for AI agents.
- Email and file protection: Secure emails, documents, and domains by reducing exposure to spoofing, phishing, weaponized files, unsafe extensions, and legacy protocols.
Get started with Secure Now
Secure Now, already in GA, provides guidance to customers on how to improve their readiness against common misconfigurations and vulnerabilities. We continue to update it with the latest guidance, recent additions include guidance on critical vulnerabilities such as PHP2Shell and threat actors' use of agentic AI tooling like ARTEX to find and abuse exposed secrets. Evaluate your organization today at aka.ms/securenow
Introducing the On-Premises Security Initiative
We are introducing a new on-premises security initiative, initially focused on server workloads, to help customers get their secure foundations across their estate. Bringing together guidance, migration, and product investments, we will help organizations understand and prioritize risk, reduce their on-premises footprint by moving workloads to Microsoft 365 wherever possible, and secure critical workloads that need to remain on-premises. Guidance for the initiative will soon be available through Secure Now. Learn more in our on-premises security initiative announcement.
Additionally, we will introduce On-Premises Baseline Security Mode, which is designed to strengthen the default security posture of supported Exchange Server, SharePoint Server, and Skype for Business Server workloads, reducing the manual effort required to establish and maintain recommended security settings while preserving administrator control.
Get started
Discover what else is new with Microsoft Perception and get started today.
Frequently Asked Questions
How can I get started with AI-readiness Initiatives?
- To use AI-readiness initiatives at preview in January, customers must have access to Perception.
Will I still be able to use initiatives in Microsoft Security Exposure Management?
- Yes. Any initiatives you currently use in Microsoft Security Exposure Management will continue to work.
How are AI Readiness Initiatives different from recommendations in Secure Now?
- AI readiness initiatives based off the same set of recommended controls across domains as Secure Now. Secure Now helps provide actionable exposure management guidance publicly, and anyone with an Entra ID can use it. AI Readiness Initiatives in Perception apply the guidance to your own environment, measure progress with live tenant data, and use agents to help do the work.
How can I get access to Perception Green Agents?
- Sign up for Perception.
Do Perception agents work with multi-cloud setups?
- Yes, they will work across Azure, AWS and GCP data connected to Defender for Cloud