Gunter Ollman explains the benefits of CPSM technology, how IT security teams have evolved, and how the pandemic has shaped security.
Organizations have experienced “two years’ worth of digital transformation in two months,” Microsoft CEO Satya Nadella said earlier this year. Much of this change has centered on cloud adoption, which has been a key part of supporting the shift to work-from-home environments.
Gunter Ollmann, chief security officer for Microsoft’s Cloud and AI Security division, saw the redesign of access to corporate assets and the cloudification of on-premises assets as more businesses decided to work remotely in the long term. While end-user security has improved throughout the acceleration, a few security gaps remain that still need to be addressed.
“The digital transformation has added a whole other layer to the environment that has to be managed and secured,” Ollmann said in an Interop Digital keynote interview with Dark Reading executive editor Kelly Jackson Higgins. “That adds a lot of new stress, a lot of new skills – and there’s a skills gap that still need to be closed – and that also translates to resources.”
This transition has worsened businesses’ existing challenges with misconfigurations and other cloud security issues, he continued. Much of this could be related to lack of related skills on the on-premise teams, which creates an additional burden for organizations to work through.
From a cloud operator’s perspective, Ollmann is seeing the growth of cloud security posture management (CSPM) technologies, which are meant to help security teams bring together their assets and resources in one place to better manage and understand their cloud infrastructure.
“CSPM has been that vehicle for providing visibility of security risk, vulnerabilities, vulnerability management, and then a little bit of gamification to enable and help customers and organizations improve their security posture as they go along,” he explained.
Security posture management gives infosec teams visibility and control while managing policies. The gamification – a “loose interpretation” of the term, Ollmann noted – is in the score, which informs teams of the risk or security value in a particular asset, resource, application, or environment as a whole. Every vulnerability and poor or absent configuration has a value tied to it. By addressing the weaknesses, a team can increase its overall security score.
“Security will never be 100%, so hopefully as you develop these sorts of things, you keep improving on your score,” he said. Some larger businesses have multiple apps in multiple environments, and teams compete against one another to boost their numbers.
To gauge the effectiveness of a scoring system, Microsoft looked at multifactor authentication (MFA). About 80% of asset compromises in the cloud are due to lack of MFA, Ollman said, so they gave it a higher score.
“As soon as we bumped up the score for MFA and that was in people’s overall security score, we suddenly started to see people turning it on and using it more,” he added.
Juggling Responsibilities: IT, Security, Networking
Networking and security teams are core to the way online businesses are developed and run. As new technology is adopted and distributed across the enterprise, networking and security professionals have changed the way they work together and support business processes.
The cloud-first environment demands a shift in skillsets, Ollmann explained.
“It really has become, infrastructure is code and configuration is code,” he said. “What it means is those assets, those resources become ephemeral, and you have to change the way that you operate.” Most of today’s security pros must have some networking expertise, and vice versa, he added.
More security and networking experts are playing the role of internal consultants for business executives, he pointed out. Their role is shifting away from saying, “We should do this, and this is why” and toward, “Let me understand your operations and how you need to business.” They’re becoming more of a peer to business leaders, who are beginning to take more ownership over security.
“That responsibility for, ‘How do I secure,’ ‘How do I configure my network,’ ‘How do I operate my business,’ now has been delegated, and that responsibility is shared across those teams inside organizations,” he said.
Even with the additional collaboration, today’s CISOs and IT operations professionals are overwhelmed with alerts, misconfigurations, missing patches, compliance issues, and other problems that demand attention, Ollmann pointed out. A goal of his is to cut down on the clutter and triage and prioritize these alerts with artificial intelligence.
Anti-spam, anti-phishing, and anti-malware tools “are pretty much overtaken by AI,” he added. Years of malware and phishing samples are distilled into new AI systems that create machine learning classifiers that are then deployed and work much better.
“They’re stopping more and more threats even though the bad guys are also investing in machine learning and AI to try to bypass these classifiers,” he said.
When it comes to defending attacks that target companies’ machine learning systems, Ollmann said a lot of research is going on in the space. Microsoft, along with a group of organizations, recently created a dictionary of techniques used to attack machine learning models. The threat matrix serves as an extension of the MITRE ATT&CK framework for classifying attack techniques.
Kelly Sheridan is the Staff Editor at Dark Reading, where she focuses on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial … View Full Bio