Tanium has been selected as a testing partner for Microsoft's Codename MDASH, placing it among a limited group in a private preview of the security system.
Microsoft describes Codename MDASH as a multi-model agentic security system that uses more than 100 specialised AI agents to identify and validate potential code vulnerabilities. Tanium is one of a small number of customers testing the system and providing feedback during its early development.
The partnership focuses on software security at a time when companies are under pressure to find flaws in their code before attackers can exploit them. The companies are concentrating on weaknesses that traditional static analysis tools can struggle to detect, including logic problems, authorisation issues and vulnerabilities spread across multiple files.
Tanium used the system against its own production codebase during the preview. It then combined MDASH's findings with AI-assisted static and dynamic validation, followed by human review, before sending confirmed issues into its secure development lifecycle.
Tanium described that process as a layered model in which automated analysis surfaces potential problems and security engineers decide which findings are valid and require action. The approach mirrors the detect, decide and remediate model Tanium uses across its wider IT and security products.
Christian Hunt, Chief of Engineering at Tanium, said the partnership reflects a broader shift in how software is built and secured.
"Our customers build their IT and security programs on Tanium, so the security of what we ship is critical for us," Hunt said.
"We quickly recognised how AI is accelerating how technology companies build and secure software, and we have been applying frontier model capabilities across our software development lifecycle. Being selected to test Codename MDASH lets us push these efforts further, alongside the companies defining what's possible. It is one more reflection of a commitment that has always been non-negotiable for us: earning the trust our customers place in Tanium every day," he said.
AI in Security
The use of artificial intelligence in code analysis has moved beyond experimentation and into operational security programmes at large organisations. As software development cycles speed up and attackers adopt more advanced methods, businesses are looking for ways to reduce the time between discovering a flaw and fixing it.
Traditional static analysis tools are widely used to review code for known patterns of weakness, but they can miss more complex vulnerabilities that do not fit simple rules. Multi-agent systems such as MDASH are being developed to tackle those harder cases by assigning different tasks to different AI models and combining the results.
For Tanium, the value of joining the preview lies not only in early access to a new tool, but also in shaping how it fits into a broader review process. Confirmed findings from the MDASH testing were routed directly into Tanium's existing engineering and security workflows rather than treated as standalone outputs.
Loic Simon, Vice President of Security at Tanium, said the company was focused on what happened after the scan.
"The value is not just running a scan, it is what you do with what it finds that matters," Simon said.
"We ran MDASH against Tanium code and put every meaningful finding through our own validation and remediation process. Combining agentic AI with the judgment of our security engineers is how we raise the bar on the software our customers depend on, and it is why we want to help shape tools like MDASH early," he said.
Broader Pressure
The announcement reflects wider pressure on software companies to strengthen internal development controls as code vulnerabilities become a more prominent attack surface. Organisations are not only defending live systems and employee devices, but also examining the security of the software they write and ship to customers.
Tanium is known for endpoint management and security software, and says it is applying the same operational discipline it offers customers to its own code. In practice, that means using machine-led detection to examine large volumes of software while leaving validation and prioritisation to experienced security staff.
Microsoft has not outlined a full public launch for Codename MDASH in the material released by Tanium, but the private preview suggests it is testing the system with selected partners before wider availability. Tanium's role gives Microsoft feedback from a security software supplier using the tool on production code rather than test material.
The collaboration also underlines how major technology groups and specialist security vendors are increasingly working together on AI-based development and defence tools. For enterprise software users, the central question is whether those tools can reduce false positives, find weaknesses earlier and fit into existing remediation processes without adding another layer of manual work.
In Tanium's account of the preview, human review remained the final checkpoint before any issue entered its development lifecycle, with security engineers responsible for proving and prioritising the findings.