DXC Technology Collaborates with Primary on Zero-Trust Security for Enterprise AI

DXC Technology Co. and security startup Primary announced a collaboration today, positioning DXC as the sole managed services partner for Primary’s zero-trust platform, which is specifically engineered for enterprise artificial intelligence. This joint solution aims to regulate how AI agents and applications access data within an organization, incorporating identity verification and access policies before systems are engaged. All requests are logged, supporting compliance requirements.

DXC will deliver this service through its global cybersecurity division, encompassing various aspects such as consultation, implementation, integration, governance, and ongoing operational support. Both companies assert that their offering addresses a gap in current enterprise security tools, which are typically designed for human users, devices, and networks, rather than autonomous AI agents.

AI agents operate on independent schedules, interact with other tools, and traverse systems that were not initially configured for self-governing software. Primary’s platform enforces attribute-based access control via a policy engine, treating users, agents, services, and devices uniformly within its identity layer. The company also tracks agent prompts, responses, and tool invocations.

Access control is enforced at the session layer, with data loss prevention mechanisms applied to all incoming and outgoing data. Furthermore, Primary provides a software development kit for teams seeking to integrate these controls directly into their custom applications.

Primary’s leadership includes co-founder and Chief Executive Greg Marx, alongside co-founder and President Michael Marx. Peiter “Mudge” Zatko, formerly the security chief at Twitter, heads Primary Labs, a unit within the company. Dawn-Marie Vaughn, DXC’s global offering lead for cybersecurity, emphasized the necessity for governance to evolve at the same pace as AI deployments, noting that trusted AI requires more than just infrastructure security.

Vaughn highlighted the importance of controlling how AI systems interact with data, identities, and business processes. Michael Marx commented on the complexity of the challenge, explaining that the difficulty lies not just in observing an agent’s actions, but in possessing sufficient context to determine the legitimacy of those actions, coupled with the infrastructure to enforce such decisions in real-time.

Initially, the two companies plan to target organizations in highly regulated sectors, where stringent data sovereignty and auditing demands are paramount. DXC’s financial results for the first quarter of fiscal year 2027 showed revenue of $3 billion, with organic revenue experiencing a 6.7% decline. Bookings, however, increased by 5%. Following the July 30 report, the company's shares decreased by approximately 9%.

During the same presentation, DXC also showcased its own AI initiatives, including an agentic security operations platform that reportedly reduced analyst handling time from 80 minutes to four, illustrating the company's broader engagement in AI-driven solutions.