On November 16, 2026, OpenAI announced a major expansion of its ChatGPT for Academic Researchers program (originally launched as a pilot on July 29, 2026) alongside a companion study by the OpenAI Economic Research team titled Science at the Frontier: How Scientists Use ChatGPT. Rather than introducing a new foundation model, the updates revise grant capacity, subscription tiers, and selection rules while sharing pilot usage findings.
Program Expansion: 200,000-Researcher Target, Nov 23 Launch & Shift to ChatGPT Plus
Since the pilot began in July 2026, 20,000 researchers across 33 countries have joined the program. Under the November 16 expansion:
Doubled Program Target: OpenAI states it is expanding the program to give up to 200,000 researchers worldwide free access to ChatGPT—doubling the 100,000-researcher goal announced in July—while the companion report refers to free access for more than 200,000 researchers globally.
Applications Open November 23 (First-Come, First-Served): The next application round opens on Monday, November 23, 2026. Replacing the pilot's lottery-based selection, OpenAI will award 50,000 new grants on a first-come, first-served basis to the first 50,000 applicants who meet eligibility requirements (qualified users already on the waitlist will receive account activation instructions by email before applications open).
ChatGPT Plus Tier & Up to Two Collaborators: Each new grant covers the recipient and up to two collaborators—supporting up to 150,000 researchers in this round—with free access to ChatGPT Plus instead of the pilot's ChatGPT Pro tier, after usage reviews showed most participants stayed within Plus usage limits. Data shared through the program is excluded from model training by default. The announcement also cross-references existing initiatives including ChatGPT Edu, GPT-Rosalind and the Rosalind Academic Plan, and OpenAI's $5 million Simons Foundation mathematics and science commitment.
Science at the Frontier Report: Pilot Usage & Self-Reported Time Savings
Using privacy-preserving classifiers without reading underlying user messages, the OpenAI Economic Research team analyzed 580,186 science-related messages from 1,819 pilot scientists between August 20 and September 22, 2026, paired with survey responses from 251 scientists collected September 30 to October 7, 2026:
Self-Reported 1.9x Time Estimate: Across eight research activities (249–251 respondents per activity, with unchanged time counted where AI was not used), surveyed scientists estimated that completing the same work at the same quality without AI would have taken 1.9 times as long.
Ideation & Discovery Expectations: 68% of scientists used ChatGPT to develop research ideas (averaging 35% scientist-led, 36% co-developed, 23% suggested by ChatGPT, and 6% other), and 98% of 250 respondents expected AI to accelerate discovery in their field (citing faster execution at 52%, finding more promising questions at 23%, and pursuing broader questions at 14% across 367 coded reasons from 224 written responses). OpenAI notes that productivity figures are self-reported, discovery findings reflect expectations rather than causal effects, and researchers continue to face staffing, funding, and experimental bottlenecks.
