DXC Technology and AI company Anthropic have announced a multi-year global partnership to deploy Anthropic’s Claude models in mission-critical enterprise systems operated by DXC for customers including banks, airlines, insurers, manufacturers and government agencies.
Under the arrangement, DXC will train “tens of thousands” of forward-deployed engineers and builders who will be certified on Claude through Anthropic’s Partner Academy and embedded in customer environments, the companies said. DXC said selected engineers will complete training and certification within 90 days and will focus on “designing, deploying, and governing agentic AI systems” in production environments.
The partnership expands on DXC’s existing use of Claude, including in the development of DXC OASIS, described by the company as an AI-native orchestration platform for managed services. DXC said Claude is the default foundation model for OASIS agentic workflows and that the platform is deployed across more than 50 customers following its April 2026 launch.
DXC said Claude helped accelerate OASIS software delivery by an estimated 10 times and that more than 95% of code was generated by Claude before human review. The companies did not provide independent verification of these figures.
Paul Smith, Anthropic’s chief commercial officer, said DXC had tested Claude in its own operations “under the same security and compliance requirements their customers face” before expanding deployments with customers. DXC president and CEO Raul Fernandez said the alliance would scale Claude “directly into the mission-critical technology systems we run for our customers”.
DXC said initial focus areas for new offerings would include insurance, cybersecurity, application services and code modernisation. In cybersecurity, the announcement referenced a “DXC OASIS security engineer sub-agent” built on “Claude Security” for use across security operations centres.
The announcement comes as large enterprises and government agencies assess how to operationalise generative AI in regulated and safety-critical environments, where data governance, model risk management and auditability remain key concerns. Embedding AI models into managed services and core infrastructure can increase productivity and automation but also expands the attack surface and heightens the need for controls around access, prompts, data leakage and model outputs.
