General Dynamics Information Technology (GDIT) and OpenAI have formed a partnership to deploy secure AI capabilities across federal agencies. The agreement aims to move government AI adoption beyond limited pilot programs and into operational workflows that handle sensitive data.
What the partnership will deliver
Under the collaboration, GDIT will integrate OpenAI's frontier models - including GPT-5.6 - into federal environments using its own mission and systems integration expertise. The companies plan to deploy AI agents that automate routine workflows, manage complex case workloads spanning multiple agencies, and analyze large data sets.
GDIT will also bring OpenAI's Codex coding agent into secure federal settings. Codex will review, test, and modernize legacy applications, including systems written in COBOL and Fortran. The work falls under GDIT's Vision, Acceleration and Innovation (VIA) strategy, which structures technology investments and commercial partnerships.
Systems integration and compliance
GDIT said the integrations will connect agencies' existing infrastructure with OpenAI models while protecting personal information and controlling data sharing. The approach is designed to meet federal compliance requirements and support the operation of sensitive government systems - a hurdle that has kept many agencies from moving AI Agents & Automation into production.
The partnership fits a broader pattern at GDIT. In June, the company expanded its alliance with Splunk to build AI-enabled cybersecurity capabilities, including agentic cyber operations and zero trust functions. That followed a March agreement with Seekr to develop agentic AI for federal missions.
What leadership is saying
Ben Gianni, senior vice president and chief technology officer of GDIT, framed the deal around a shift in how government work gets done. He said the partnership is meant to help agencies move toward a future in which "human expertise and autonomous AI agents work together to deliver public services."
Joseph Larson, OpenAI's vice president and head of government, said the collaboration "brings Codex's advanced reasoning and OpenAI's secure API delivery models to federal missions," giving public sector developers tools to modernize infrastructure and deliver results faster.
Why this matters for government professionals
This partnership signals that the infrastructure for secure, compliant AI for Government is maturing beyond the pilot stage. For federal IT and program managers, the near-term impact will be the availability of AI agents that can operate inside existing agency systems - handling cross-agency casework, legacy code modernization, and large-scale data analysis - without breaking compliance boundaries. The question shifts from "can we use this?" to "how fast can we train teams to work alongside autonomous agents?"
Your membership also unlocks:
