The question of "who owns artificial intelligence" is far more complex than simply identifying the creator of an AI system. It delves into the evolving legal landscape surrounding intellectual property, the unique structure of AI organizations like OpenAI, and the very nature of AI's autonomous capabilities.
The concept of artificial intelligence ownership is multifaceted, encompassing the AI models themselves, the data used to train them, and the output they produce. Traditional intellectual property (IP) laws, primarily designed for human creativity, are grappling with the rapid advancements of AI technologies.
The field of artificial intelligence is seeing an unprecedented boom, with new AI products and AI startups emerging constantly. This rapid development creates significant challenges for existing legal frameworks.
Copyright and Generative AI: One of the most debated aspects is the ownership of content generated by AI. In the US, for instance, the Copyright Office has generally determined that AI-generated works without significant human authorship do not qualify for copyright protection. This means that if an AI system, based solely on a human prompt, determines the expressive elements of its output (text, images, etc.), that output typically cannot be copyrighted. However, if a human provides substantial creative input to the AI's output, they may claim copyright over the enhanced work. The UK, on the other hand, has seen conflicting views, with some arguing copyright should belong to users and others advocating for exclusion from ownership entirely.
Patent Law and AI: When it comes to inventions created with or by AI, patent law becomes more relevant. However, the UK Supreme Court, in the 2024 DABUS case, upheld rulings that AI-generated inventions cannot be patented, as inventorship is traditionally attributed to human creators. This highlights the ongoing challenge of applying existing patent law to autonomous AI systems. The key question is whether an AI innovation represents genuine creativity or is simply an output of its programming and algorithms.
AI Training Data: A crucial element of artificial intelligence is the data used to train AI models. The ownership of this AI training data is a significant issue, especially when copyrighted work is used without explicit permission. Companies like Getty Images have initiated legal battles against AI companies, alleging unauthorized use of their copyrighted material for AI training. Clear legal agreements defining data rights and responsibilities are becoming essential to prevent disputes in AI projects. Privacy regulations like GDPR also impact data ownership by requiring explicit consent and anonymization for personal data used in AI training.
OpenAI, a prominent player in the artificial intelligence space, particularly with its large language model ChatGPT, has a unique ownership structure that reflects the complexities of AI development.
Hybrid Structure: OpenAI was initially founded as a non-profit in 2015 by figures like Elon Musk and Sam Altman, with a goal to promote and develop friendly AI for the benefit of all humanity. However, it later transitioned to a "capped-profit" model, with a non-profit entity overseeing a for-profit arm (OpenAI LP).
Control and Investment: The original OpenAI Nonprofit retains control, with its board of directors setting targets and strategy for the for-profit entity. Microsoft has made significant investments in OpenAI, including over $11 billion, and reportedly has a 49% stake in the profits of OpenAI's for-profit unit. However, Microsoft does not own the non-profit or have a controlling majority stake in the entire organization. Other institutional investors also contribute, but the non-profit's board maintains ultimate authority.
The global AI market is highly competitive, with countries and tech companies vying for dominance. This is evident in the patent landscape.
Global Patent Filings: Over 340,000 AI-related patents were filed globally between 2010 and 2020, with an annual growth rate of 28% between 2015 and 2020.
Country Leadership:
China leads in terms of sheer volume, accounting for nearly 70% of all global AI patents. In 2024, China accumulated roughly 70% of the world's AI patents, with approximately 300,000 applications. Chinese inventors filed 38,210 generative AI patents between 2014 and 2023.
The United States ranks second, with around 18% of global AI patent filings. While filing fewer patents than China (around 67,800 applications in 2024), US AI patents are cited nearly seven times more often, suggesting a focus on quality and impact. In 2022, the US granted approximately 12,077 AI patents, about 21% of all global AI grants that year.
Europe holds approximately 5% of AI patent filings worldwide.
Japan and South Korea together contribute about 6% of AI patents.
Industry Adoption: AI patents now represent about 10% of all technology-related patents filed annually. In terms of adoption, 78% of global companies report using AI in their business as of 2025. This includes 71% using generative AI in at least one business function. Larger enterprises are twice as likely to use AI than smaller businesses.
The question of AI ownership will continue to be a central debate as artificial general intelligence (AGI) and more autonomous systems develop. MIT Technology Review and other leading publications regularly highlight the ethical considerations, legal ambiguities, and societal implications of breakthroughs in artificial intelligence.
The development of AI capabilities also raises concerns about AI safety and the responsible AI development. The discussions extend to who is accountable for the output of AI, especially in cases of bias or errors from an AI system. As the market for AI grows, projected to reach $1.85 trillion by 2030, balancing innovation with clear frameworks for ownership and accountability will be paramount for the entire AI ecosystem.
