Microsoft has unveiled a new inline data protection feature for its Edge for Business browser, designed to prevent sensitive corporate information from being shared with consumer AI applications like ChatGPT, Google Gemini, and DeepSeek.

“With the new inline protection capability for Edge for Business, you can prevent data leakage across the various ways that users interact with sensitive data in the browser, including typing of text directly into a web application or generative AI prompt,” Microsoft explained.

The company plans to expand this protection to cover additional generative AI tools, email platforms, collaboration software, and social media applications in the future.

## Enhanced Teams Security

Alongside this browser update, Microsoft announced the general availability of collaboration security features for Microsoft Teams, addressing recent phishing attacks conducted by threat actors like Storm-1674 and Storm-1811, who have used Teams to distribute malware and deploy ransomware.

The new security controls allow organizations to:
– Specify which tenants, domains, and users can communicate with employees
– Protect against malicious links and attachments through real-time analysis
– Improve reporting mechanisms for suspicious messages

Microsoft’s security approach includes a sandbox environment where “suspicious files and URLs are automatically executed in a secure, isolated environment to determine if they exhibit any malicious behavior.” This real-time detonation process identifies and neutralizes harmful content before users can access it.

## Security Copilot Expansion

Microsoft is also expanding its Security Copilot with 11 new AI agents, including five from external partners. These agents will help organizations analyze data breaches, prioritize alerts, perform root cause analysis, and improve compliance.

The Microsoft-developed agents, available for preview next month, will:
– Triage phishing alerts
– Monitor data loss prevention and insider risks
– Track vulnerabilities and remediation efforts
– Curate threat intelligence based on organizational exposure

“The relentless pace and complexity of cyber attacks have surpassed human capacity and establishing AI agents is a necessity for modern security,” said Vasu Jakkal, corporate vice president at Microsoft Security. “The phishing triage agent in Security Copilot being unveiled today can handle routine phishing alerts and attacks, freeing up human defenders to focus on more complex threats and proactive security measures.”