DXC already runs the IT infrastructure behind some of the world's most regulated industries. Now it is embedding Claude inside those systems — and training an army of certified engineers to make it stick.
Introduction
Most enterprises that want to adopt AI face a familiar obstacle: the technology looks compelling in demonstrations but hitting production in a regulated, compliance-heavy environment is a different matter entirely. Banks cannot experiment with AI in the middle of core transaction systems. Airlines cannot take risks with operations software. Government agencies cannot deploy AI in ways that sidestep the security and audit requirements their systems are built around.
On June 11, 2026, Anthropic and DXC Technology announced a multi-year global alliance designed specifically for this reality. DXC — one of the world's largest IT services companies, with 115,000 employees operating across 70 countries — already runs the mission-critical infrastructure these industries depend on. The alliance will put Claude inside those systems, backed by tens of thousands of DXC engineers trained and certified specifically for that work.
Quick Summary
| Detail | Info |
|---|---|
| Partnership Type | Multi-year global alliance |
| Announced | June 11, 2026 |
| DXC Size | 115,000 employees · 70 countries · 50+ years in IT services |
| FDE Program | Tens of thousands of Claude-certified forward-deployed engineers |
| Certification Path | Anthropic Academy + DXC's own mission-critical curriculum |
| OASIS Launch | April 2026 · 50+ customers · 95%+ code written by Claude |
| Development Speed | Claude sped up DXC's software development by 10x |
| Initial Focus Areas | Insurance · Legacy modernization · Cybersecurity · Application services |
Who DXC Is — and Why That Matters Here
DXC Technology is not a typical software vendor. For more than 50 years — across its current form and the companies it was built from — DXC has operated the backend technology infrastructure that large institutions cannot afford to have fail. Its client base includes some of the world's largest banks, major airlines, leading insurers, global manufacturers, and national government agencies.
These are not organizations experimenting with AI in a sandbox. The systems DXC manages handle real financial transactions, insurance claims, flight operations, and government services at scale — under strict regulatory, security, and compliance requirements that leave very little room for error.
That institutional context is the foundation of this announcement. Bringing AI into these environments is not about plugging in an API. It requires engineers who understand the compliance requirements, the legacy architecture, and the operational stakes. It requires AI that has already been validated in those same conditions before being handed to a client. And it requires a deployment model that works within the governance frameworks these industries have built over decades.
The Anthropic-DXC alliance is structured around all three.
DXC Proved the Concept in Its Own House First
The most credibility-bearing detail in this announcement is not the scope of the partnership. It is the sequence.
Before bringing Claude to a single client, DXC deployed it inside its own operations. The company has 115,000 employees working under the same kind of security and compliance requirements its customers demand. Claude was used internally at that scale first — giving DXC the ability to say, with direct operational evidence, that the deployment works in conditions equivalent to what its clients will require.
That internal validation produced one of the more striking numbers in the announcement. DXC used Claude as its primary tool to build DXC OASIS — its new AI-native orchestration platform for managed services, launched in April 2026. Claude wrote more than 95% of the OASIS codebase. Engineers reviewed and validated that code, but the generation itself was AI-driven. The result: software development ran ten times faster than it would have under traditional approaches.
OASIS now runs actively for more than 50 DXC customers, with a global rollout underway. Claude is the default foundation model powering its agentic workflows — not a feature, but the operating core.
The Forward-Deployed Engineer Program
The centerpiece of the alliance for long-term scale is the forward-deployed engineer (FDE) program. DXC will train tens of thousands of its engineers to become Claude-certified — not as general AI users, but as specialists embedded directly inside client organizations to implement and operate Claude within their specific systems.
FDEs are a different model from traditional consulting. They do not advise from the outside and hand over a report. They work inside the client environment, shoulder to shoulder with the client's own teams, on the actual systems. That embedded model is particularly suited to the industries DXC serves, where implementation decisions are inseparable from deep familiarity with the underlying infrastructure.
Certification runs through Anthropic Academy — Anthropic's training and certification program for engineers working in its partner ecosystem. DXC has built an additional curriculum layer on top of the Anthropic Academy foundation, covering the mission-critical system environments specific to DXC's client base. An engineer certified through DXC's program is not just Claude-literate — they are trained for the specific complexity of deploying AI in regulated, high-stakes infrastructure.
Four Areas Where the Work Starts
The alliance launches with focus on four domains where DXC already operates at meaningful scale and where Claude can contribute useful work today:
Insurance
DXC will deploy Claude to help insurance clients modernize their core systems and build agentic solutions. The approach is tailored: each deployment accounts for the client's specific business context, operating model, and strategic priorities rather than applying a generic template. Insurance systems carry decades of accumulated architecture and regulatory requirements — the work of modernizing them requires both deep AI capability and domain familiarity.
Modernization as a Service (MaaS)
Legacy codebases are one of the most expensive and risk-laden challenges in enterprise IT. Millions of lines of code written in older languages, without documentation, by engineers who have long since moved on — analyzing, refactoring, and modernizing these systems traditionally takes years and carries high failure rates. DXC is using Claude to compress that timeline and improve accuracy, offering legacy modernization as a service to enterprise clients who need to move faster than traditional methods allow.
Cybersecurity
For OASIS, DXC is building an always-on security engineer subagent powered by Claude Security — Anthropic's security-focused AI capability. This subagent will be deployed across DXC's security operations centers (SOCs), providing continuous monitoring and response capabilities for the organizations whose infrastructure DXC manages. An AI agent that never clocks out, constantly watching for threats and anomalies across mission-critical systems, is a meaningful shift in what a SOC can realistically maintain.
Application Services
Beyond security, DXC is embedding Claude directly into the application maintenance and management environments it operates for clients. This means AI working inside the operational fabric of the applications themselves — not a separate tool a human consults, but an agent integrated into the day-to-day work of keeping enterprise applications running.
The Claude Partner Network
DXC is also now a member of the Claude Partner Network — Anthropic's formal network of consulting and services firms that deploy Claude in enterprise environments. The network represents Anthropic's strategy for reaching large organizations through partners with existing deep relationships and technical infrastructure, rather than exclusively through direct sales.
For DXC, membership in the partner network formalizes an already operational relationship. For Anthropic, DXC's scale — 115,000 employees, 70 countries, 50 years of enterprise relationships — represents a distribution channel that reaches into industries where Anthropic would not easily gain entry on its own.
What Makes This Different From a Standard AI Partnership
Enterprise AI partnerships are announced constantly. Most follow a recognizable pattern: a technology vendor agrees to embed another company's AI into its products, both companies issue a press release, and the actual deployment happens slowly — if at all — over the following years.
The Anthropic-DXC announcement has a few features that separate it from that pattern.
DXC is already in production. OASIS launched in April 2026 and serves 50+ customers. The partnership is scaling something that exists, not announcing something that might be built.
The internal proof-of-concept is documented. DXC built OASIS using Claude, 95% of the code, at 10x the usual development speed. That is not a benchmark number — it is an operational result from inside DXC's own organization.
The FDE model is structural. Training tens of thousands of engineers to be Claude-certified and embedding them inside client organizations is not a pilot program. It is an infrastructure investment that, once made, becomes a durable capability DXC owns.
The target industries are the hardest ones. Banks, airlines, and government agencies are not early adopters. Getting AI into the systems these organizations depend on is the kind of enterprise adoption that typically takes years and generates significant revenue when it works.
What Has Not Been Disclosed
The announcement is high on scope and intentionally light on financials. Several details remain undisclosed:
- The total financial value of the multi-year alliance
- The exact timeline for training "tens of thousands" of FDEs
- Specific client names across banking, aviation, insurance, or government
- The full global rollout schedule for OASIS
- Detailed terms of the Claude Partner Network membership
Organizations watching this space should treat the disclosed numbers — 95% of code, 10x development speed, 50+ OASIS customers — as the most concrete signal of what the partnership has actually delivered so far.
Final Takeaway
The promise of AI in regulated industries has been discussed for years. What has been missing is not capability — it is the combination of technical credibility, institutional trust, and deployment infrastructure needed to move from discussion to production inside systems that cannot fail.
DXC brings 50 years of institutional relationships, operating experience in the most demanding compliance environments on earth, and a platform — OASIS — that was already built on Claude and is already running for paying customers. Anthropic brings the AI. The alliance is structured around the one thing both companies know these industries actually need before they commit: proof that it works in their conditions, from people who have already done it themselves.