OpenAI-GSA OneGov Deal Extends ChatGPT Access to 2028

What happened

OpenAI has renewed its OneGov agreement with the US General Services Administration, extending ChatGPT's availability to federal agencies through 2028. The updated deal marks a shift from the original flat-rate licensing structure to a consumption-based pricing model, according to FedScoop.

The renewal builds on OpenAI's earlier OneGov arrangement with the GSA, which had opened the door for US government bodies to access ChatGPT under negotiated terms. The multi-year extension signals continued federal appetite for generative AI tools despite the shift to usage-based costs, which typically ties spend more directly to actual adoption across agencies.

Why it matters

The renewal is significant less for its length than for its pricing structure. Moving from flat-rate to consumption-based billing changes the economics of government AI adoption: agencies pay in proportion to actual usage rather than a fixed subscription, which can lower the barrier to initial trials but also means costs scale with success. For a technology vendor, it is a bet that usage will grow steadily across a large, historically slow-moving buyer base.

For public-sector technology leaders, the deal reinforces that generative AI is moving from pilot to sustained infrastructure inside government. A multi-year horizon gives agencies more confidence to build workflows, training and procurement processes around ChatGPT rather than treating it as a short-term experiment — a meaningful marker for GovTech modernisation more broadly.

The Renascence take

Pricing model changes rarely make headlines, but they quietly shape how technology actually gets used inside large, risk-averse organisations. A shift to consumption-based billing is as much a behavioural lever as a commercial one.

Most coverage will focus on the "through 2028" headline, but the real story is the pricing switch. Flat-rate deals encourage broad, low-friction adoption because cost is predictable regardless of use; consumption-based pricing rewards agencies that build real workflows around the tool but can also discourage experimentation if usage costs aren't well understood upfront. Any public-sector body signing up under this model should invest early in usage governance and internal reporting — not just to control spend, but to build the evidence base needed to justify further AI investment to budget holders who will be watching consumption curves closely.