The recent success of DeepSeek, a generative artificial intelligence system released under an open-source license, has deeply shaken the world of technology and the software industry. One of the most surprising reactions was the statement by Sam Altman, CEO of OpenAI, who publicly admitted that he had been ‘on the wrong side of history’: “I personally think we have been on the wrong side of history here and need to figure out a different open source strategy; not everyone at OpenAI shares this view, and it’s also not our current highest priority.”

The success of open-source AI, why now?

This statement, as rare as it is significant in the industry, reopens the debate on the business model of software development companies. For decades, the industry has seen the confrontation between the proprietary model, characterized by closed-source code and paid-for licenses, and the open-source model, focused on community sharing and collaboration. The proprietary business model originated with the software industry and has long been the dominant standard. In this approach, the source code is closed and only accessible to the developer company, which retains total control over the distribution, use and modifications of the software.

Historical examples of proprietary software include Microsoft Windows, Adobe Photoshop and VMware. Companies following this model generate revenue through the sale of user licenses, with an initial acquisition cost. Support and maintenance fees are often required annually for updates and service. Subscription models (Software as a Service – SaaS) guarantee a constant revenue stream over time. This model has guaranteed solid profits for large companies for years but has several disadvantages, including high costs for users, especially for companies requiring multiple licenses. Vendor dependence (vendor lock-in) limits the flexibility of companies. Less innovation, as development is limited to the capabilities and vision of the developer company.

The rise of open source: from pioneering to maturity

Open-source software, on the contrary, is based on a completely different paradigm. The source code is open, accessible and modifiable by anyone, fostering collective innovation and the rapid evolution of technologies. An emblematic example is Linux, the operating system that revolutionized the world of servers and embedded devices. Initially, open source was a niche phenomenon reserved for technicians and enthusiasts, but over time, it has gained maturity and credibility thanks to the emergence of Open-Source distributions, such as Red Hat and Ubuntu. Companies like Red Hat have created a sustainable business model by offering 24/7 support and professional technical assistance. Ready-to-use distributions, eliminating the need to manually compile source code.

Hybrid solutions, combining free software with value-added services. This model has enabled Open Source to enter the enterprise market of Companies and Public Institutions and compete directly with proprietary software, offering advantages such as zero acquisition costs, with payment only for support. Greater transparency and security, thanks to open and verifiable code. Flexibility and independence, avoiding lock-in to a single supplier.

The open-source artificial intelligence revolution

Until now, generative artificial intelligence seemed destined to follow the proprietary model. Platforms such as ChatGPT and Gemini were developed as closed products. However, the arrival of Deepseek, an Open-Source model developed in China, has overturned this logic. Its free release put US companies under pressure, causing an earthquake in the financial markets: the shares of OpenAI, Microsoft and other AI giants fluctuated sharply while the Open-Source model quickly gained acceptance. Why did DeepSeek change the rules of the game? Free and modifiable access, allowing developers around the world to customize and improve it. Reduced costs for companies that can integrate the model without paying high licenses. Acceleration of innovation, thanks to a global community working on continuous improvement. DeepSeek’s success proves that open source is not only a viable alternative but also a faster and more efficient driver of innovation than the closed model.

DeepSeek open-source

Already, US company Hugging Face is racing to replicate R1 with OpenR1, a clone of DeepSeek’s model that Hugging Face hopes will reveal even more of the ingredients of R1’s special sauce, its efficiency-enhancing innovations. The impact of the Open-Source revolution will not be limited to the world of AI. Other areas of information technology will also undergo profound transformations. One example is the data centre market, which has been dominated for years by proprietary software such as VMware, recently acquired by Broadcom. Open-source alternatives such as Red Hat (USA) and Vates (Europe) are emerging as increasingly attractive options for companies due to lower costs and greater flexibility. This trend could lead to a redistribution of power within the technology industry, reducing the dominance of large proprietary companies.

2025, the year of the AI open-sourced

We are facing a momentous turning point in the software industry. Sam Altman’s statement is just one sign of a much bigger change that will see more and more companies and developers embrace Open Source as a winning business model. The year 2025 could be remembered as the year of Open Source, the moment when the technology industry finally recognized the value of accessibility, collaboration and transparency. The consequences will be profound and lasting, and we will witness a new era of digital innovation, more open and democratic than ever before.