OpenAI Founded
In December 2015, a group of technology investors and researchers (including Greg Brockman, Ilya Sutskever, Wojciech Zaremba, John Schulman, Elon Musk, and Sam Altman) announced the founding of OpenAI, a non-profit artificial intelligence research laboratory in San Francisco, with approximately one billion US dollars in pledged funding.
Background
By 2015, the organisations doing serious AI research could be counted without much trouble. Google had acquired DeepMind in 2014, bringing one of the most ambitious research groups in the field under a commercial roof. Facebook had its own AI Research lab. Microsoft, IBM and a handful of others were building teams. The pattern was clear: frontier research was moving inside large technology companies, where it sat alongside product lines, revenue targets and shareholder expectations.
That concentration worried some people in the field. The concern was not that the research would stop, but that it would narrow. Safety questions, long-term risks, and work that did not obviously lead to a product could all lose out when a commercial organisation had to choose where to spend money. Publishing findings openly was another pressure point. A company with a competitive advantage in a technique had every reason to keep it quiet.
There was no obvious institution to hold a different position. Academic labs existed, but could not attract or keep researchers at the salaries the large companies were offering. The gap between what a university could fund and what a well-resourced private lab could do had grown considerably as the cost of training capable models climbed.
What happened
In December 2015, a group of researchers and investors announced the founding of OpenAI, a non-profit research laboratory based in San Francisco. The announcement named Sam Altman, who was president of Y Combinator at the time, and Elon Musk, then known primarily for Tesla and SpaceX, alongside researchers including Ilya Sutskever (who had been at Google Brain), Greg Brockman, Wojciech Zaremba, John Schulman, Andrej Karpathy, Pieter Abbeel, Durk Kingma, Vicki Cheung, Jonathan Ho and Jonathan Ziegler. The backers pledged about one billion US dollars in funding, though the announcement made clear this would be drawn on over time rather than handed over at once.
The structure was deliberate. A non-profit with no obligation to return money to shareholders could, in principle, work on problems that were important rather than problems that were profitable. The founding blog post said the organisation’s goal was to advance digital intelligence in the way most likely to benefit humanity as a whole. Research findings would be published and shared openly, rather than kept inside the organisation. That was the stated intention, at least.
The founding team was a striking mix. Some came from academic machine learning, others from industry research groups, and several would later become well-known in their own right. Sutskever had co-created AlexNet with Geoffrey Hinton and Alex Krizhevsky in 2012, the convolutional network that had pushed image recognition accuracy well beyond what previous methods could achieve. Having someone of that standing choose a new non-profit over an established commercial lab made people pay attention.
OpenAI did not ship any product in its early months. It was a research organisation starting from scratch, and the work it was committing to, particularly in reinforcement learning and language modelling, was long-horizon. What existed in December 2015 was a founding document, a team, and a funding pledge.
Why it mattered
OpenAI's founding introduced a well-capitalised research organisation explicitly structured around safety and open publication at a time when frontier AI research was increasingly concentrated inside a small number of large technology companies. Its non-profit structure and stated mission to develop AI for the benefit of humanity rather than shareholders represented a deliberate institutional counterpoint to commercial laboratories. The organisation subsequently published influential work in reinforcement learning, language modelling, and generative AI, making it a significant actor in the trajectory of the field.
People
Sam Altman, Greg Brockman, Ilya Sutskever, Wojciech Zaremba, John Schulman, Elon Musk, Pieter Abbeel, Vicki Cheung, Andrej Karpathy, Diederik P. Kingma, Jonathan Ho, Jonathan Ziegler
Organisations
OpenAI, Y Combinator, Tesla, Deepmind, Google
Sources
- Introducing OpenAI.OpenAI.Primary source
- Non-profit OpenAI launches with $1B pledge from Musk, Thiel and others.TechCrunch.Secondary
- Elon Musk and Others Drop $1 Billion to Launch OpenAI.Wired.Secondary
Cite this page
AI Achievements. (2015). OpenAI Founded. Retrieved 2026-08-22, from https://achievements.ai/milestone/openai-launched-by-elon-musk
@misc{achievements_openai_launched_by_elon_musk,
title = {OpenAI Founded},
author = {{AI Achievements}},
year = {2015},
url = {https://achievements.ai/milestone/openai-launched-by-elon-musk}
}