DXC Technology Lands Exclusive Deal for AI Agent Identity Management Platform

DXC Secures Exclusive Channel for AI Agent Access Control

DXC Technology announced an exclusive managed services partnership with Primary on August 6, 2026, positioning itself as the sole enterprise provider of Primary's AI-native Zero Trust platform. The service targets identity and access management for AI agents and enterprise AI applications, combining data access controls, security policy enforcement, and compliance management under a single managed offering.

The deal matters because it shifts AI agent identity from a roadmap item to a packaged service available through an existing enterprise vendor relationship. Organizations already using DXC for security outsourcing can add AI agent IAM as an incremental line item rather than evaluating and deploying new standalone products. For DXC's existing customer base, this creates a path to policy-driven access control for AI workloads without standing up separate infrastructure or hiring specialized staff.

What Changes for Enterprise Buyers

The partnership creates immediate pressure on incumbent IAM vendors to demonstrate explicit controls for non-human identities. Microsoft Entra, Okta, CyberArk, and SailPoint all address machine identity to varying degrees, but Primary's positioning as "AI-native" rather than adapted from human IAM workflows forces a product differentiation question: can existing platforms handle the volume, velocity, and policy complexity of autonomous AI agents making thousands of data access decisions per minute, or do they require bolt-on modules and workarounds?

For CISOs writing AI security RFPs in late 2026, the DXC-Primary service sets a baseline capability: treat AI agents as first-class identities with policy-driven data access, not ad hoc API keys or credentials embedded in application code. Organizations deploying LLM-based applications, autonomous workflows, or AI agents with access to sensitive data now face a clear vendor question—does your IAM stack manage AI agent identities natively, or are you patching together point tools?

The announcement lacks pricing detail, deployment metrics, or customer references, which limits direct economic comparison. DXC's model shifts budget from software licenses to managed service contracts, a structure that favors organizations with constrained security headcount but may cost more over multi-year terms than self-managed platforms.

Market Growth Validates Expanded IAM Budgets

Straits Research published updated IAM market forecasts on August 13, sizing the global market at $24.56 billion in 2026, up from $21.52 billion in 2025, and projecting growth to $70.66 billion by 2034. The 14.12% compound annual growth rate confirms IAM as one of the fastest-growing security categories, justifying larger budget allocations in multi-year plans.

For procurement teams, this trajectory means vendors will push aggressive product expansion—AI-driven governance, privilege management for machine identities, and tighter integration between IAM and security operations platforms. Expect both list price increases and bundled discount offers as vendors compete for larger deals in a growing market. Organizations consolidating IAM platforms or expanding from single sign-on to full identity governance can position these projects as aligned with market momentum rather than discretionary.

Competitive Pressure on Incumbents and Startups

The DXC-Primary deal sits at the intersection of three overlapping markets: traditional IAM, Zero Trust network access, and AI security. Microsoft launched Zero Trust guidance for AI agents at RSA 2026, Okta released AI-powered identity threat protection in June, and startups like Oasis Security (which raised $120 million for "Agentic Access Management") and Oak (which raised $60 million for an "AI-native identity operating system") are positioning explicitly around non-human identity.

For buyers, this competition creates both opportunity and risk. More vendor options and feature velocity mean better negotiating leverage and faster innovation cycles. It also means overlapping terminology, unclear product boundaries, and integration complexity. A vendor claiming "AI-native" identity may deliver purpose-built agent access controls, or may rebrand existing machine identity features with AI marketing.

The key evaluation question: does the platform manage AI agent data access as a distinct identity type with its own policy model, or does it treat agents as service accounts with bolted-on monitoring? The former enables granular, dynamic access control at the scale AI workloads require. The latter creates operational friction and eventual security gaps.

What to Watch

Track DXC customer deployments and case studies in Q4 2026 and Q1 2027. If Primary's platform gains traction in regulated industries—financial services, healthcare, government—expect rapid product announcements from Microsoft, Okta, and CyberArk positioning their own AI agent identity capabilities. If adoption stalls, it signals the market is not yet ready to pay for dedicated AI identity management, and buyers can defer investment.

Monitor whether other global integrators—Accenture, Kyndryl, TCS—strike similar exclusive or semi-exclusive partnerships with emerging AI security vendors. If they do, it confirms that managed services, not standalone software, will dominate AI identity deployments for large enterprises.

For organizations deploying AI agents in production, the baseline question is now clear: can you demonstrate policy-driven access control for every AI agent, with auditable logs and real-time policy enforcement? If not, expect board-level questions and regulatory scrutiny as AI risk moves from theoretical to operational.