Microsoft’s Project Perception Gives Partners New Security AI Tools

‘Some of the partners will have security operations for customers, but they haven’t yet taken the opportunity to take the learnings from security operations and implement them as fixes in the system. And now they have the tools to do that end to end, and they have the people to do that end to end,’ says Scott Woodgate, Microsoft Security general manager.

Microsoft plans to bring its Project Perception agentic security systems into private preview Aug. 3, giving cybersecurity professionals a new way to leverage artificial intelligence against attackers who themselves are attacking more often and in new ways through AI.

For security solution providers in the Redmond, Wash.-based vendor’s 500,000-plus member ecosystem, Project Perception offers a way to go beyond Security Operations Center management to remediate more issues faster, especially key at a time when partners can’t hire security professionals fast enough, Scott Woodgate, Microsoft Security general manager, told CRN in an interview.

Woodgate added that Project Perception has the potential to increase margins for security partners through AI agents that act as members of a red team, blue team or green team—terms for security team members who simulate attackers, defenders and remediation, respectively.

“The opportunity for partners has never been bigger,” Woodgate said. “Some of the partners will have security operations for customers, but they haven’t yet taken the opportunity to take the learnings from security operations and implement them as fixes in the system. And now they have the tools to do that end to end, and they have the people to do that end to end. … There’s an opportunity to keep customers safe and grow.”

[RELATED: Microsoft’s Nicole Dezen Unveils New Partner Strategy Built Around AI Transformation]

Microsoft Eyes Aug. 3 For Project Perception Private Preview

Woodgate and other Microsoft security executives spoke with CRN during a launch event for Project Perception held in San Francisco Monday.

Kelly Yeh, president of Chantilly, Va.-based Microsoft solution provider Phalanx Technology Group, told CRN in a recent interview that his company has seen more demand for training around security and user endpoints, using not just tools from Microsoft but third-party technology as well.

Yeh has worked with his clients on mandating security training and validating that users have gone through the training, with failing to complete security modules going against employee key performance indicators for raises and bonuses if the client wants.

Helping clients quantify mandatory staff training and policy enforcement with phishing tests and other security reports can help clients with insurance, he said.

“You can give them metrics, and having that—one, it can help lower the insurance costs—but two, if you do get compromised, it’s going to make the policy enforceable. It doesn’t give the insurance company an out.”

How Project Perception Uses AI Agents To Detect And Remediate Threats

While early leaks of Project Perception focused on its capabilities that can unearth vulnerabilities, the breadth of the system actually shows Microsoft’s prowess as a security vendor, Woodgate said.

Project Perception includes purpose-built AI models including Microsoft’s MAI-Cyber-1, organizational context, full-estate signals and sensors, orchestration harness and actuators that help AI agents act on a decision, according to the vendor. Project Perception is meant to work with Microsoft’s generative AI-assisted Security Copilot chat interface.

Like Security Copilot, Project Perception is consumption-based, pay-as-you-go pricing measured in Microsoft Security Compute Units. Agents don’t consume SCUs at the same rate, with task intensity a factor in the rate, according to Microsoft.

Microsoft Defender users will have access to Project Perception’s multi-agent defense from within the tool, with Project Perception extending to other Microsoft security products over time, according to the vendor.

Aleš Holeček—a 22-plus-year Microsoft veteran promoted in February to Microsoft Security president and chief architect—told CRN in an interview at the event that Project Perception can help security teams augment their existing skill sets while also automating away lower-level concerns to focus on bigger security challenges.

“Our target with Perception is to kind of help [security] people to live a little more normal lives,” Holeček said. “It is a very difficult and challenging landscape environment to be to be in 24x7.”

Microsoft is the vendor to bet on for security because of its massive install base equipped with security signal sensors that can also conduct remediation, plus an intelligence layer in the middle to help manage security context in real time, he said.

Microsoft Promises Partners Agentic Security Revenue Opportunities

The event comes as Microsoft solution providers continue to digest the slew of new partner promotions and benefits the vendor has rolled out in its new fiscal year, which started July 1. Microsoft reports earnings for the first quarter of the new fiscal year Wednesday.

During the recent Microsoft Customer and Partner Solutions (MCAPS) Start for Partners digital event by the vendor, Aarti Borkar, corporate vice president of customer success for Microsoft’s $20 billion-plus security business, said that ecosystem partners can resolve 60 percent to 70 percent of cybersecurity incidents automatically, growing revenue without increasing head count.

Large, enterprise security partners leaning into services with intellectual property, white labeling and co-managed offerings are seeing 20 percent year-on-year revenue growth. Managed services providers are growing at 24 percent. In small and midsize partners, they are seeing 23 percent overall growth year on year and 42 percent growth in managed services.

“And we expect even more upside in the age of agentic security because these numbers still don’t account for the agentic opportunity, which will only expand as the addressable market keeps expanding,” she said.

At the end of July, Microsoft is moving to an audit-based model for its four security specializations—cloud security, data security, identity and access management, and threat protection, according to the vendor.

Microsoft is also designing an agentic security specialization for partners who use AI to detect, investigate and respond to threats at machine speed, Steve Thomas, director of the Microsoft AI Cloud Partner Program (MAICPP) offers and benefits team, said during the event.

Microsoft has also increased its security post-sales investment in partners four times this fiscal year, with the goal of helping them assess customer risk and build security strategies, Microsoft Partner Incentives Senior Director Uriel Rootshtain said during MCAPS.

The vendor increased the amount partners can earn from deployment accelerator engagements by 30 percent, introduced a conversion bonus for verified displacements of select competitors and now allows partners to earn up to 80 percent more per Microsoft Sentinel engagement, with payouts when initial deployment is completed and when consumption goals are achieved, Rootshtain said.