$42.6 billion "Challenge" in US dollars? OpenAI seeks to use equity to loosen regulatory restrictions.

Against the backdrop of soaring valuations in the artificial intelligence (AI) industry and increasing political pressure, a game revolving around the distribution of shares among AI giants and regulatory boundaries is unfolding in Washington. According to reports, OpenAI has proposed to transfer 5% of its shares to the US government as an attempt to alleviate pressure from the political arena. However, while President Trump released subtle signals regarding AI regulation on Thursday, he avoided discussing the issue of shares. Meanwhile, sources familiar with competitor Anthropic revealed that the company has not discussed government investment with the Trump administration. OpenAI extends an "olive branch" of shares According to two sources cited by the media, OpenAI CEO Sam Altman proposed the idea of holding 5% of shares in early discussions with the Trump administration. Based on OpenAI's post-financing valuation of $85.2 billion in March of this year, these shares would be worth approximately $42.6 billion. Altman argues that allowing the public to have a financial interest in the company is the best way to share the benefits of AI. This proposal extends beyond just OpenAI. It is reported that the idea is a broader arrangement, where through a government entity, Washington would hold 5% of shares in every leading AI developer in the US. The potential targets mentioned include Anthropic, Alphabet Inc. Class C (GOOGL.US), and Meta (META.US). According to the plan, these shares would be held by the government through a sovereign wealth fund instrument. However, it is currently unclear whether the mentioned companies are willing to accept OpenAI's proposal. The White House, OpenAI, Alphabet Inc. Class C, and Meta all did not respond to requests for comments on the matter. It is understood that OpenAI proposed the creation of a "public wealth fund" in April of this year to hold assets capturing the growth of AI companies and distribute economic benefits to the public. The preliminary negotiations regarding the 5% shares are a continuation of this idea. Anthropic denies discussions on shares, export control tensions easing Regarding speculations about whether other AI companies are also conducting similar discussions on share allocations, a source familiar with the matter stated on Thursday that the Trump administration has not had any discussions with Anthropic regarding government investment. Anthropic declined to comment on this. In fact, Anthropic had just experienced friction with the US government over controls on advanced model exports. Last month, the US Department of Commerce imposed restrictions on Anthropic's most advanced Mythos 5 and Fable 5 models citing cybersecurity concerns, requiring the company to obtain government approval before opening them up to any foreign users. This forced Anthropic to temporarily suspend the use of the models and swiftly engage in discussions with US officials on security concerns. Last week, the government began gradually easing the restrictions, allowing approved US entities to regain access, and this week, export controls on Fable 5 have been fully lifted. Although the restrictions were ultimately lifted, the initial decision by the US government sent shockwaves through Silicon Valley, seen by outsiders as an indication of a shift in the Trump administration's traditionally lenient stance on AI regulation. Concerns are rising in Washington about the possible misuse of advanced AI models by military intelligence agencies of countries like Russia, leading to a noticeable increase in the scrutiny from regulators on the release of new models, although the submission for review currently remains voluntary. Trump: Need guardrails but not too many, dodges share issue Facing escalating industrial and political pressure, Trump stated on Thursday that it is necessary to establish certain standards for AI, but regulation must not be too strict. "We do need some regulatory guardrails, but fewer the better." He added, "When we think there are players not playing by the rules, posing a certain risk, we block them quickly and effectively." However, he did not specify any specific targets. When directly asked about the reports of OpenAI negotiating to provide 5% of shares to the government, Trump avoided the question and instead discussed the government's investment in Intel Corporation (INTC.US) in 2025. He often emphasizes that the deal generated additional income for taxpayers. In August of last year, the US government acquired 10% of Intel Corporation's shares after investing $8.9 billion in common stock; in May of this year, Trump even stated that they should have demanded a larger stake at that time. He described the government holding shares in AI giants as a "wonderful thing" and would make Americans "partners in this revolution." However, the debate over the distribution of wealth in AI goes beyond just share proposals. In recent months, as concerns grow about AI replacing human jobs, calls for sharing benefits have been increasing. Senator Bernie Sanders has even proposed using the tax system to capture wealth driven by AI, requesting large AI companies to relinquish 50% of their shares to the government and give government board seats. This proposal is much more radical than the 5% suggested by OpenAI. Currently, companies like OpenAI and Anthropic are preparing for their initial public offerings, and government investment schemes, export control boundaries, and public statements by political figures are collectively outlining a complex picture of the US AI industry seeking a difficult balance between innovation, regulation, and wealth distribution. The fluctuating signals from policies, coupled with the proactive release of share intentions by companies, make the final outcome of this game still full of variables.