OpenAI has, on its own merits, become one of the most prominent reference companies worldwide. Until recently, it was a relatively unknown company; however, in recent years it has achieved significant recognition thanks to one of its most pivotal tools: ChatGPT. (Could you even imagine our personal and professional lives without ChatGPT?)
For this reason, we have decided to provide a comprehensive review of the history of this artificial intelligence development company, from its beginnings with Sam Altman—and yes, Elon Musk as well—up to the present day.
OpenAI was founded in 2015 by Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman, Trevor Blackwell, Vicki Cheung, Andrej Karpathy, Durk Kingma, John Schulman, Pamela Vagata, and Wojciech Zaremba. This diverse group of entrepreneurs, developers, and artificial intelligence specialists would lay the foundation of what the company would become in the future.
Additionally, Jessica Livingston, Peter Thiel, Reid Hoffman, and the companies Amazon Web Services, Infosys, and YC Research acted as donors (along with Sam Altman and Elon Musk themselves), contributing one billion dollars. The objective of this investment was to form a team of AI experts in order to develop the project safely and with the aim of transforming it into a useful tool for everyone.
Initially, OpenAI began as a non-profit organization and was created with the intention of developing advances in AI in a safe manner and with meaningful societal impact. Its spirit was to foster collaboration among professionals, as well as promote the publication of all work conducted—making such knowledge accessible to the world.
In February 2018, Elon Musk resigned from OpenAI’s board of directors. According to certain media outlets, his departure from the project was due to the refusal by Altman and the rest of OpenAI’s board to allow Musk to assume control of the company.
Several years later, in 2024, Musk would go on to file a lawsuit against OpenAI and its CEO, Sam Altman, before the San Francisco Superior Court alleging breach of contract. The entrepreneur argued that OpenAI had shifted to the pursuit of self-interested economic benefit, deviating from its original objective of achieving a public, open-source general AI.
However, 2018 was also the year of the arrival of the first GPT model (Generative Pre-trained Transformer) launched by OpenAI, which marked a milestone in the development of language models based on transformer architecture—focusing on pretraining neural networks to understand context and generate coherent text.
To better understand this concept, imagine you have a long sentence and you wish to comprehend it word by word, but also how each word relates to the others to give the entire sentence proper meaning. Transformer architecture works in a similar way, enabling machines to understand sequences of information such as text, code, or even data. Unlike earlier models that processed information word by word, a transformer can view the entire sequence simultaneously, enabling it to discern relationships between words that are distant from each other within the sentence.
GPT-1 was regarded as an innovative model for its time, as it was trained in an unsupervised manner using vast amounts of text to learn language patterns and generate high-quality text. Although its impact was not as significant as that of its successors, it laid the foundation for the development of more advanced models such as GPT-2, GPT-3, and eventually GPT-4.
Currently, although OpenAI continues to be governed by a non-profit organization, since 2019 it has maintained a unique capped-profit model. Furthermore, it is the non-profit organization, through its board of directors, that governs and supervises the activities that generate profits.
This model consisted in the creation of a for-profit entity within the non-profit organization: a for-profit subsidiary was established—OpenAI LP (Limited Partnership)—to oversee the majority of research and development activities and to commercialize AI technologies. The non-profit organization (OpenAI Inc.) would continue to govern the LP through its board and would set a maximum cap on the financial returns that investors could receive, as a multiple of their initial investment.
In any event, the principal objective remained to benefit humanity and to develop AI safely.
In 2021, OpenAI continued expanding its influence in the field of artificial intelligence with the popularization of GPT-3, a language model that was remarkable for its ability to coherently and naturally generate text, closely resembling human language. GPT-3 saw widespread adoption across various applications, from content creation to the development of virtual assistants. This advancement cemented OpenAI’s position as a leader in generative AI research and development.
In that same year, Microsoft deepened its relationship with OpenAI. The Redmond-based company had invested one billion dollars in OpenAI in 2019, but in 2021, Microsoft announced a more extensive collaboration. OpenAI utilized Microsoft’s Azure infrastructure to train its models, which enabled OpenAI to improve the capacity and services of its AI models.
Undoubtedly, however, 2021 will be remembered for the launch of Dall-E, an AI model that generated images from textual descriptions. This demonstrated the potential of AI to understand and create visual content with a high level of creativity and caused a tremendous social impact, marking the starting point for the era of image generation from text.
The first ChatGPT was launched by OpenAI in November 2022. This model was based on GPT-3.5, an improved version of GPT-3, and was designed to engage users in conversational interactions. ChatGPT represented a significant advancement, as it enabled generating responses that were coherent and contextually relevant in a chatbot format, thus rendering it accessible and useful for a wide range of applications—from personal assistance to content generation and customer support.
The launch of ChatGPT attracted extensive media attention and evolved into a viral phenomenon owing to its ability to sustain high-quality conversations in natural language. Within just a few days, thousands of users began interacting with the model, which prompted OpenAI to make adjustments and improvements, expanding access more broadly to the community.
Furthermore, OpenAI subsequently launched a subscription plan called ChatGPT Plus in 2023, which enabled users to access the most advanced version, GPT-4, featuring significant enhancements in performance and reasoning capacity.
OpenAI’s established position in the generative AI market was shaken in late 2023 by a series of internal changes, which began with the dismissal of Sam Altman (CEO) and other executives from the company on November 17.
Three days later, Emmett Shear was appointed as OpenAI’s new CEO. Shear is a co-founder and former CEO of Twitch, and has a more tempered view toward AI advancement compared to Altman. Nevertheless, pressure from investors (including Microsoft) and from the majority of OpenAI employees proved successful, and Sam Altman was reinstated as CEO on November 22.
The year 2023 was also notable for the arrival of GPT-4. Featuring substantial improvements over GPT-3 and GPT-3.5, GPT-4 was introduced as possessing a far greater capability to understand and generate responses in more complex and technical contexts.
In October 2024, OpenAI achieved a historic milestone by closing what was, at that time, the largest funding round in history. The company raised an extraordinary 6.6 billion dollars (approximately €5.98 billion). This catapulted OpenAI’s valuation to 157 billion dollars (around €142 billion), placing it well ahead of other AI companies.
The main investor in the operation was Thrive Capital, which invested approximately 1.3 billion dollars (€1.178 billion), with an exclusive option to invest up to an additional 1 billion dollars at the same valuation (through 2025). Microsoft, Nvidia, SoftBank, Khosla Ventures, Altimeter Capital, Fidelity, and MGX also participated as investors.
These companies reportedly set the condition that OpenAI’s corporate structure be changed within the next two years in order to remove the cap on economic returns that investors could receive. This would entail a transition from a non-profit entity to a for-profit organization.
Nevertheless, OpenAI stated that “The non-profit organization is fundamental to our mission and will remain in existence.”
By 2025, OpenAI had grown to the point where it had become a company capable of generating 10 billion dollars in annual recurring revenue (ARR) derived from product and service licensing.
These revenue figures add context to the company’s valuation of 40 billion dollars following its March 2025 funding round. The 40-billion-dollar funding round (approximately €37 billion), led by Japan’s SoftBank (an investor in companies such as Uber and Alibaba), became the largest in history. This elevated the company’s valuation to 300 billion dollars, positioning OpenAI as the third-largest unicorn globally at that time.
A few days earlier, OpenAI debuted its new image generation model within GPT-4o, enabling the creation of more realistic works, a higher level of detail, the inclusion of coherent text, and enhanced consistency for generating variations, among other useful capabilities. The result? The fervor for “Ghibli-style” images.
A few weeks earlier, one of the first measures taken by Donald Trump after assuming the office of President of the United States was to announce the Stargate project. This public-private partnership seeks to position the United States as a global epicenter for AI. OpenAI is one of the technology companies participating in the project, along with Microsoft, Nvidia, Oracle, and Arm.
Stargate contemplates an estimated investment of 500 billion dollars (approximately €479.175 billion) in artificial intelligence infrastructure over four years. The primary objective is to build a network of advanced data centers and to promote the development of practical AI applications in key areas (health, security, and the economy).
However, it cannot be said that 2025 was an entirely positive year for OpenAI, as in December the company had to activate a “red code” due to the course its sector positioning had taken. Amid a confusion of products and models—straddling the worlds of enterprise solutions and mass-market AI, and without a clear path to profitability—Sam Altman’s company set a new course to prevent competitors such as Google or Anthropic from capturing its market share.
Indeed, once again. In 2026, the company announced, yet again, the largest funding round in history. 122 billion dollars, which increased the company’s valuation to 852 billion dollars, consolidating its position as the second most valued startup in the world. This funding round was led by existing investors such as Amazon, NVIDIA, and SoftBank, as well as Microsoft and other investment groups from around the globe.
This funding round was accompanied by a clear declaration of intent that closely relates to the strategic change announced at the end of 2025. Following the cancellation of its video application, Sora, OpenAI announced it would seek to restructure its product offerings, unifying them under a single “superapp”. While this is still forthcoming, we may for now consider the company’s extraordinary numbers: it is already reaching two billion dollars in monthly revenue, and its pilot advertising program has already achieved more than 100 million dollars in ARR in under six weeks since inception.
What will be OpenAI’s next significant move? Will we ultimately see it achieve artificial general intelligence?
Photo: Depositphotos
Your email address will not be published. Required fields are marked *
Δ