Vladimir Lenin’s observation that “there are decades where nothing happens, and there are weeks where decades happen” aptly describes our current moment in artificial intelligence development. The release of DeepSeek R1, developed in China at a fraction of the cost of comparable models, represents more than just technological advancement. It marks a fundamental shift in who can develop and deploy AI systems. This democratization of AI technology will present challenges for lawmakers at all levels of government.
What is DeepSeek?
DeepSeek R1 is an AI chatbot, much like Open AI’s ChatGPT. But DeepSeek R1 has several differences from ChatGPT. First, the code for DeepSeek’s reasoning model is open source, making it easy for any aspiring AI developer to download and build applications on it. Because DeepSeek released the code for free, unlike models from large American tech companies, there is a lower cost barrier to access this model.
In addition, DeepSeek’s founders announced that the model was built for less than $6 million, a claim that spooked the U.S. stock markets because American tech companies are spending hundreds of millions more on their models. Some observers doubt that DeepSeek is accurately reporting its true training costs, with one independent research report estimating that the true cost was more than $500 million. In any case, it is clear that AI development is no longer limited to a handful of American tech companies; DeepSeek may be the first inexpensive open-source AI reasoning model, but it is not the last. The stark reality is that AI development can no longer be contained within the walls of well-resourced tech companies, representing a double-edged sword for social progress.
The Benefits of AI Democratization
The democratization of AI could make open-source models like DeepSeek a “profound gift to the world,” as tech investor Marc Andreessen proclaimed. Indeed, DeepSeek promises to level the playing field by enabling small businesses and developing nations to compete in AI development without massive computing infrastructure.
For example, small to mid-size businesses now will not need to build their own computing models or purchase licenses from large tech companies. These cost savings will give broad access to advanced computing to businesses and individuals, while likely forcing other AI companies to bring costs down. More people will be able to design bespoke AI applications for their businesses; as tech giant IBM noted, this accessibility will push innovation forward. In this way, democratization could foster healthy competition and innovation that benefits both businesses and consumers.
Moreover, AI development is currently concentrated in a “handful of technology mega-corporations.” A 2017 study found that “only around 10,000 people in roughly seven countries [were] writing the code for all of AI.” A recent Stanford study revealed that U.S. developers produced 61 models in 2023, while the EU created 21 and China produced only 15. Hence, most of the world is excluded from the opportunity to participate in AI development, an issue that this author knows firsthand from participation in the United Nations Development Programme’s Discussion Group on AI and Development in Latin America and the Caribbean. DeepSeek’s arrival disrupts this concentration.
The democratization of AI will create reduced barriers to entry and add more unique voices and solutions to the AI ecosystem. Local developers will gain the ability to adapt the technology to address specific regional needs. Broader accessibility will lead to more localized and culturally relevant AI applications. It will also reduce algorithmic bias, since a more diverse set of developers can identify and correct biases that might otherwise go unnoticed.
Finally, open-source models may be generally viewed as more trustworthy than corporate models. The World Economic Forum notes that offering users the chance to “interrogate training data” engenders more trust. Transparency allows developers to join together to review methodologies and address security flaws in the design and application of open-source software. However, while increased transparency and accessibility drive innovation, they also create new risks as AI development tools become available to those who may use them irresponsibly or maliciously.
The Drawbacks of AI Democratization
Democratization raises significant concerns about responsible AI development and oversight. While established players like Microsoft AI and Anthropic have demonstrated commitment to social responsibility and risk mitigation, smaller actors and businesses in regions with limited regulatory frameworks may not adhere to the same standards.
One major issue is data privacy. This is already an issue for lawyers, who must be concerned about uploading confidential information to any AI tool. But open-source AI chatbots present a sui generis concern: What happens to proprietary information users upload to a chatbot? How, if at all, is a user’s own data being protected? Data privacy concerns have already led to multiple governments banning DeepSeek, including Australia and Italy. Legislation is currently pending in the U.S. House of Representatives to ban DeepSeek on government devices, and individual states (including New York) are also barring government officials from downloading DeepSeek onto their work devices.
Data privacy is not the only concern. Open-source AI chatbots in the hands of bad actors can create scary scenarios, such as developing bioweapons, promoting self-harm among teenagers, and facilitating the spread of mis- and disinformation. Environmental risks are also present, as the challenges of tracking energy consumption and preventing misuse by bad actors become more complex as AI development becomes more distributed. These risks demand thoughtful regulatory responses.
What’s a Lawmaker to Do?
Rather than pursuing outright bans, which may simply push development underground and beyond oversight, policymakers should consider more nuanced approaches. There are a few steps governments and businesses can take right now to mitigate the risks of open-source AI models.
First, policymakers must begin by educating themselves in a neutral way about this rapidly changing environment. Academia should play a role; this author has trained over 1,000 government officials about emerging tech and related legal issues through USF Law’s Center for Law, Tech, and Social Good.
Once they understand the issues, federal, state, and local governments should establish clear guidelines for data privacy and security. For example, California’s new slew of AI-related privacy legislation is a good start. Lawmakers can also create incentive structures that reward responsible innovation, such as financial support for small businesses to develop responsible AI applications.
These efforts are most likely to start locally, as global cooperation to quickly address these concerns seems far-fetched at the moment. Educational incentives to expand AI literacy worldwide could ultimately benefit AI model developers, particularly by cultivating access to broader markets and showing how models can be used in a wide variety of use cases. It would also serve to reduce opportunity inequality around the world. However, the lack of cooperation at the 2025 Paris AI Action Summit indicates that international standards for AI development and deployment are still years away.
Consequently, this situation also offers a crucial opportunity for the investment community to play a vital role. By directing capital toward companies committed to ethical AI development, investors can ensure that democratization and responsible innovation advance hand in hand. Recent research from VentureESG highlights how targeted investment strategies can promote both AI innovation and responsible development practices.
As a Business and Tech Law professor studying the intersection of emerging technologies and regulation, I see striking parallels between today’s AI landscape and the early days of blockchain technology. Just as blockchain’s decentralized nature disrupted traditional financial systems, DeepSeek’s open-source low-cost approach challenges the concentration of AI development among a handful of major technology companies. While this shift promises to expand access to AI capabilities, it also raises critical questions about safety, accountability, and responsible development. To properly address these concerns, lawmakers must begin by educating themselves about the opportunities and risks inherent in this pivotal moment.