As discussions about TikTok’s uncertain status in the United States continue, elements within Congress are also deliberating on the prohibition of DeepSeek, a rival to ChatGPT from China. Just last week, this application surpassed ChatGPT in the App Store rankings.

A new legislative proposal put forth by Senator Josh Hawley, a Republican from Missouri, seeks to “prevent individuals in the United States from enhancing artificial intelligence capabilities within the People’s Republic of China.” Essentially, this could place a ban on DeepSeek, even though the proposed law doesn’t specifically cite the AI chatbot.

If enacted, the proposed law stipulates that those found guilty of “importing or exporting artificial intelligence technology to or from China” could be sentenced to 20 years in prison, in addition to hefty fines—up to $1 million for individuals and $10 million for companies.

Does this mean casual users of DeepSeek would face legal repercussions? While the bill does allow for such possibilities, it is more likely that the legislation targets developers and distributors. If passed, users can expect to see DeepSeek and other Chinese AI applications vanish from app stores.

The broad wording of this bill may stem from DeepSeek’s open-source architecture, which currently allows developers to easily download and integrate its models into their applications. The law aims to limit this practice, but it could also jeopardize future international collaborations, including the ability to access Chinese research on AI.

In a statement, Senator Hawley mentioned, “Every dollar and every gigabyte of data that flows into Chinese AI translates to funds and information that could ultimately be weaponized against the United States.”

Concerns regarding DeepSeek’s implications are not new. The AI has already been banned from use by the Navy due to various “security and ethical considerations” and is also prohibited on government devices in Texas. Furthermore, OpenAI has pointed out to The Financial Times that there is evidence suggesting DeepSeek may have been developed using aspects of OpenAI’s models. While this practice is not uncommon in AI development, it violates OpenAI’s terms of service once the resulting products are disseminated. Additionally, there has already been a notable data breach, and users of DeepSeek are required to consent to keystroke tracking.

However, the stringent language within the proposed legislation raises alarms among experts regarding the future of AI innovation and oversight, as it could hinder collaboration and transparency. Kit Walsh, the Director of AI and Access-to-Knowledge Legal Projects at the Electronic Frontier Foundation, expressed to 404 Media that “This proposal endangers the growth and sharing of AI advancements within the United States. The government has previously argued that simply disseminating information online may be regarded as an export, and an interpretation of this law in that manner could reinforce the dominance of proprietary AI over open-source or academic research.” In other words, American tech companies might be inclined to adopt even more exclusive approaches regarding their AI methodologies, impacting more than just foreign competitors.