Cognizant AI chief says threat to IT services firms 'overblown'

Cognizant's chief AI officer Babak Hodjat said concerns that major new AI players will displace large IT services firms are "overblown," asserting that enterprise customers continue to require integrators and deployment help to put generative AI into production. Hodjat made the remarks on Feb. 26, 2026, framing them as a counterpoint to narratives that foresee wholesale outsourcing of enterprise AI work to model vendors.

The comment lands as technology vendors roll out increasingly capable platforms and foundation models that promise plug-and-play capabilities for corporate users. Hodjat acknowledged disruption from model and platform providers but said the reality of enterprise adoption remains rooted in complex systems work: data plumbing, custom fine-tuning, security hardening, regulatory compliance and operationalization at scale. Those tasks, he argued, preserve demand for established services firms that manage multi-vendor environments across legacy systems and cloud providers.

For policymakers and corporate procurement leaders, Hodjat's assessment reframes where the risk lies: not in the immediate obsolescence of integrators, but in the quality and governance of AI deployments. Large organizations must still translate models into business processes, build monitoring and rollback mechanisms, and meet sector-specific regulatory and privacy requirements. That set of activities tends to favor firms with deep integration capabilities and existing relationships inside regulated industries such as finance, healthcare and government contracting.

Institutionally, the comments underscore a bifurcated market. On one side are model and platform vendors selling increasingly standardized capabilities; on the other are systems integrators proposing tailored solutions that weave models into enterprise workflows. The competitive pressure is real, and margins on commoditized work have compressed. But the integrator role, covering customization, systems integration, change management and ongoing operations, remains acute for many clients, particularly those with critical infrastructure or strict compliance obligations.

The outlook carries implications for workforce policy and corporate strategy. If large services firms retain a meaningful share of deployment work, national workforce programs that aim to reskill tech workers for AI engineering and MLOps will need to align with employer demand for systems-level expertise. At the same time, companies and regulators face pressure to clarify procurement rules, liability frameworks and certification standards for AI systems so that integrators and vendors alike understand accountability for harms or failures.

For investors and corporate boards, Hodjat's framing suggests a transitional period rather than an abrupt market reordering. Services firms will need to accelerate investments in automation, proprietary industry solutions and partnerships with model providers to protect contracts and margins. They will also have to demonstrate governance capabilities that enterprise clients and regulators can point to when evaluating risk.

Hodjat's "overblown" characterization does not remove the strategic urgency for integrators. Rather, it shifts the emphasis: the threat is not total displacement but sustained disruption that will reward firms able to combine platform access with execution, compliance and long-term operational support. How well firms and public institutions adapt to that reality will shape who captures the next phase of enterprise AI spending.