OpenAI Economic Research Exchange: API Access Opens for Academic Labour Impact Studies

OpenAI Opens Data Pipeline to Academic Economists Studying AI's Labour Impact

LONDON, 10 June 2026 — OpenAI has launched the Economic Research Exchange (ERX), a structured programme granting academic economists access to de-identified ChatGPT usage data and frontier AI model capabilities, enabling independent empirical research into artificial intelligence's measurable effects on employment, wages, and productivity.

The programme, announced via the company's official research page, represents the most systematic effort by a major AI developer to make proprietary usage data available for peer-reviewed economic inquiry. Applications are open to researchers at accredited universities and research institutions globally.

What Happened

The ERX provides two primary resources: access to anonymised, aggregated logs from ChatGPT interactions, and compute credits enabling researchers to run experiments on OpenAI's models. Approved applicants gain access through a structured data-sharing agreement that sets privacy and publication protocols.

OpenAI has identified five research priority areas: labour market displacement and augmentation; wage effects across occupational skill bands; business productivity in enterprise deployments; inequality in access to AI tools across income levels; and macroeconomic multiplier effects of widespread AI adoption.

Participants must submit proposals to an independent review panel, commit to open publication of findings, and agree to data-handling terms consistent with OpenAI's privacy policy.

Key Facts

• Programme name: Economic Research Exchange (ERX), operated by OpenAI with independent review oversight
• Eligible applicants: academic economists and researchers at accredited institutions worldwide
• Data provided: de-identified, aggregated ChatGPT interaction logs and API compute access
• Research priority areas include labour displacement, wage effects, enterprise productivity, and AI access inequality
• All accepted research must be published openly and submitted for peer review

Why It Matters

The launch addresses a persistent gap in AI economics research: analysts and policymakers have debated AI's employment effects largely without access to first-party usage data at scale. Studies from the National Bureau of Economic Research and the Brookings Institution have relied on proxy measures — job posting rates, productivity surveys, earnings reports — rather than direct observation of AI adoption patterns.

The ERX changes that by providing economists with actual interaction-level data, enabling causal inference studies that could settle contested questions about whether AI substitutes or complements human workers across occupational categories. Early findings from the programme are expected to inform labour policy discussions in the United States, European Union, and G7 economies, where regulators have sought empirical grounding for AI governance frameworks.

Enterprise buyers evaluating AI adoption strategies may also benefit: rigorous productivity research using real ChatGPT data provides a more credible baseline for return-on-investment modelling than vendor-sponsored case studies.

What Happens Next

OpenAI has indicated a rolling applications process, with the first cohort of accepted researchers expected to begin data access by the third quarter of 2026. The company will publish a summary of accepted research projects, though individual datasets and methodology details remain subject to the signed data agreements. Independent economists at institutions including MIT, Stanford, and the London School of Economics have been cited among the programme's early academic advisers.

The ERX arrives as competing AI developers face increasing legislative pressure in Europe and North America to demonstrate transparency over economic externalities. OpenAI's move sets a precedent that rivals including Google DeepMind, Anthropic, and Meta AI may face pressure to match.

Conclusion

The Economic Research Exchange is the most substantive step any frontier AI developer has taken to open proprietary usage data to independent economic scrutiny. Its findings could shape AI governance, enterprise adoption strategies, and labour policy for the remainder of the decade.

Disclosure: BUSINESS 2.0 has no commercial relationship with companies mentioned.