Microsoft Security Blog: Microsoft unveils Microsoft Security Copilot agents and new protections for AI

Mar 24, 2025

—

Source URL: https://www.microsoft.com/en-us/security/blog/2025/03/24/microsoft-unveils-microsoft-security-copilot-agents-and-new-protections-for-ai/
Source: Microsoft Security Blog
Title: Microsoft unveils Microsoft Security Copilot agents and new protections for AI

Feedly Summary: Learn about the upcoming availability of Microsoft Security Copilot agents and other new offerings for a more secure AI future.
The post Microsoft unveils Microsoft Security Copilot agents and new protections for AI appeared first on Microsoft Security Blog.

AI Summary and Description: Yes

Summary: The text discusses Microsoft’s advancements in security measures leveraging AI, particularly through the introduction of AI agents within their Security Copilot platform. These innovations aim to enhance security operations by autonomously managing security tasks, addressing the increasing complexity of cyber threats like phishing, and ensuring governance and protection over AI technologies.

Detailed Description:

– **AI-Driven Security Solutions**: Microsoft is enhancing its Security Copilot with AI agents tailored to address various security challenges, emphasizing the need for an autonomous response to cyber threats due to rising attack volumes.

– **Identification of Threats**: Phishing attacks are highlighted as a significant concern, with Microsoft detecting over 30 billion phishing emails. The new agents are designed to triage these alerts effectively, allowing human security teams to concentrate on more complex issues.

– **Integration of AI Agents**: Six new agents from Microsoft and five from partners are being incorporated into Security Copilot to address different security needs:
– **Phishing Triage Agent**: Manages phishing alerts, improving detection accuracy.
– **Alert Triage Agents**: Streamlines data loss prevention tasks and prioritizes insider risk alerts.
– **Conditional Access Optimization Agent**: Monitors user and app registrations to identify security gaps.
– **Vulnerability Remediation Agent**: Handles app vulnerabilities for more efficient patch management.
– **Threat Intelligence Briefing Agent**: Provides curated threat intelligence tailored to organizational risk profiles.

– **Focus on Privacy and Compliance**: The emergence of the Privacy Breach Response Agent signifies a push towards compliance with growing regulatory demands, enabling organizations to respond to data breaches more effectively.

– **Innovations in AI Governance**: The text stresses the importance of securing AI as organizations increasingly adopt generative AI technologies. New capabilities aim to manage the security posture across various AI models and platforms, addressing vulnerabilities and ensuring compliance.

– **Multimodal AI Security Posture Management**: Microsoft Defender is widening its coverage to enhance security posture management for AI across various platforms like Azure and Google Cloud, indicating a growing recognition of the multicloud environment’s complexity.

– **Emerging AI Threat Detection**: Upcoming features will focus on new risks identified by OWASP related to generative AI, including indirect prompt injection and sensitive data exposure.

– **Preventing Shadow AI Risks**: New controls are introduced to manage unauthorized AI app usage, which has led to increased sensitive data vulnerabilities, through the implementation of web category filters and data loss prevention techniques.

– **Future Outlook**: Microsoft positions itself at the forefront of AI security, asserting that ongoing AI research and its Security Copilot innovations are designed to elevate organizational security capabilities in an evolving cyber landscape.

In summary, Microsoft’s comprehensive strategy blends innovative AI applications with proactive security measures, addressing the complexity of modern cybersecurity challenges while ensuring compliance and protecting sensitive data across platforms. These developments hold critical implications for professionals focused on AI, cloud, and infrastructure security.

2 2025 24 3 4 5 a access accuracy Act advancement advancements agent agents AGI AI AI applications AI governance ai model AI models AI risks AI security AI technologies alerts and Application applications Arch art as attack attacks Auto autonomous availability Azure being breach breach response breaches by C capabilities CERN challenges Cloud cloud environment co complexity compliance Condi conditional access control controls Copilot coverage critical cross cyber cyber threat cyber threats cybersecurit Cybersecurity cybersecurity challenges D data data breach Data breaches data exposure data loss data loss prevention de Defender demand design detection development driven e effective efficient email end Entra environment event exp feature features first focused for front future future outlook g Gen generative Generative AI GIS Go Google Google Cloud governance gs H high Highlight HR http HTTPS human human security iCloud implementation implications in indirect prompt injection infrastructure infrastructure security injection innovation Innovations integration Intel intelligence Iron IRS J k l land led Li low man management media Micro Microsoft Microsoft Defender Microsoft Security Microsoft Security Blog Microsoft Security Copilot ML modal Mode model models Modern Monitor multi multicloud Multimodal N no o of off on operation opilot OPM opt optimization organization organizational security organizations ory out Outlook over OWASP Patch Patch Management phi phishing phishing attack Phishing Attacks phishing emails platform platforms post posture management pre prevention privacy proactive proactive security proactive security measures professionals prompt protection R rag rate RCE red regulatory regulatory demands remediation research response Risk risks Ro RoT s search sec secure security security challenges Security Copilot security gaps security measure security measures security operations security posture security posture management security solutions security teams self sensitive data SHA Shadow AI side Sig solutions source SSE Strategy T Task tasks Teams tech techniques technologies text the threat threat detection threat intel threat intelligence threats to Tor TP up US usage use user uth V vulnerabilities vulnerability vulnerability remediation web Wi x