Doubling Down on Sora 2: Wagering Regulatory Schemes for AI

Photo by Growtika on Unsplash

Sora 2 operates as the world’s largest film studio—but entirely off the books. Without owning the footage, hiring the artists, or securing the rights, it generates scenes with cinematic realism, transporting users from U.S. soil to the lands of Westeros. OpenAI has not disclosed the specific video sources that trained it, claiming Sora 2 “was trained on diverse datasets,” including copyrighted work, raising copyright infringement concerns, especially as users cannot distinguish what makes these videos fake.

While Sora 2 itself does not release copyrighted content, its generated outputs closely mimic copyrighted characters and styles, potentially creating unauthorized derivative works or outputs functionally indistinguishable from such derivative works—thereby posing copyright infringement risks and harming the licensing market. These risks were exacerbated by OpenAI’s initial policy, which allowed the use of copyrighted characters unless rightholders opted out. Although OpenAI now requires prior permission for specific uses in response to backlash against the company’s post-hoc opt-out policy, the burden remains placed on rights holders to identify their works, opt out, and monitor outputs, rather than on the model developer to secure licenses for use. As such, copyright issues remain, while Sora 2 enables many of its users (including lay consumers) to create generated videos of copyrighted content. This issue is not unique to the U.S., as OpenAI’s policies and practices raise global concerns about copyright infringement. 

Despite this opacity, OpenAI is thriving in a market backed by the U.S. government as if it cannot lose. Sam Altman claims that if OpenAI “screws up,” it should fail. Yet, policy measures such as President Trump’s tariff on AI chips and Congress’s GAIN AI Act have created structural advantages for U.S.-based AI developers like OpenAI. These tariffs restrict foreign access to cutting-edge AI chips, while the GAIN AI Act prioritizes domestic buyers and tightens export restrictions on “countries of concern.” Together, these measures have bolstered investor confidence, even though the legal foundation for Sora 2’s training remains unsettled. 

Sora 2 is not simply creating an innovative product; it is a company wagering on the U.S. government’s regulatory scheme. While the government reacts to Sora’s rise on the AI market, investors are gambling their investments on the presumption that courts will ultimately validate unlicensed training data as fair use, or that national-security politics will shield frontier-model companies from the full consequences if courts do not. Capital markets price AI as if the legal foundation has already been resolved, placing trillion-dollar bets on uncertain legal ground.

U.S.’s Response: President Trump’s Targeted Tariff and the GAIN AI Act

While investors review disclosures and await copyright settlements to make educated investment decisions, President Trump has enforced a tariff withholding advanced AI chips, such as those sold by Nvidia, from China’s AI market. Nvidia, considered the “moat” by AI experts, commits to advancing AI infrastructure by annually releasing new GPU technology. The Company’s dominance in the global AI chip market has drawn attention from China’s competing AI industry. Chips are ammunition to AI innovation and productivity. Without specialized chips exclusive to the U.S. market, investors may shy away from investing in U.S. AI companies. President Trump’s tariffs are positioned to support domestic investment and safeguard new AI advancements, thereby advancing U.S. AI development. 

Shortly after the chip tariff, Congress enacted the GAIN AI Act, requiring companies to obtain licenses to export advanced AI chips to China. The Act ensures that U.S. persons are first-in-line buyers of powerful AI chips and restricts sales to countries of concern. By certifying that U.S. persons have priority in purchasing the AI chips, the Act inherently promotes national security. Congress supports this goal, describing the Act as particularly critical for U.S. AI innovation as a “market-based approach intended to simultaneously address serious national security concerns.” Moreover, the AI stock boom serves as early evidence of the economic impact of the chip tariff and the GAIN AI Act. Nonetheless, China continues to enact its own AI regulations, while the power source of its past AI models is now exclusive in part to U.S. models. 

President Trump’s tariff restrictions on AI chips and the GAIN AI Act are positioned to prioritize the U.S.’s access to advanced AI hardware. Having first access to cutting-edge chips could translate into faster training speeds, higher-quality outputs, and continuous model improvements of U.S.-based products, making them more desirable and competitive in the global market. However, the same policies can limit U.S. companies’ access to global hardware, thereby increasing development costs. Additionally, while the GAIN AI Act can help U.S. companies, it has the potential to adversely restrict competition, limiting exports and complicating supply logistics, both of which can slow innovation cycles and make operations more expensive. 

While Sora 2 stands to benefit from increased U.S. chip availability and improved model performance, OpenAI must carefully balance its innovation with potential cost hikes to ensure that its products are not hampered by protectionist regulation. 

Direct Effects of Trump’s Tariffs on Sora 2’s Position in the AI Market 

Generally, a reduction of foreign competition stimulates U.S. market growth. In Sora 2’s benefit, decreased reliance on unstable international supply chains reduces geopolitical risks, thereby increasing investors’ trust in the U.S.’s domestic AI market. However, overly protectionist restrictions may undermine U.S. competitiveness and additionally deter international collaboration. Limited global cooperation risks politicizing hardware supply and increasing costs of innovation, thereby having the opposite effect of actually lowering investors’ confidence in the industry. Given the fine line between positive and negative protectionism, these measures favor Sora 2, a company aligned with the U.S.’s commitment to domestic AI innovation. 

Tariffs and domestic manufacturing incentives can influence companies like Nvidia to expand their U.S.-based production facilities; prioritizing U.S. markets may incentivize technologically advanced chips to enter the US innovation market. This more substantial presence helps establish a more resilient AI hardware ecosystem in the U.S., theoretically reducing dependency on foreign semiconductors and providing long-term costs once infrastructure finally scales. All this, if it were to come into fruition, can significantly support the growth of intensive models such as Sora 2.   

The above benefits depend on whether supply chain localization can survive the initial increase in costs and limited access to specialized components sourced abroad. Additionally, this assumes that the U.S. market could overcome retaliatory measures from other countries to tighten the global chip market. As such, while on the one hand, Sora 2 can benefit from a more reliable, government-backed chip supply network to improve its product and maintain leadership in AI media, its short-term costs can be detrimental. 

Overall, these policies promise stronger innovation and growth in the U.S. AI market. However, Sora 2 stands to lose by the constraints of these protectionist policies as well. As such, how OpenAI navigates these tradeoffs will determine whether Sora 2’s advantage becomes sustainable, allowing it to continue leadership in the industry. 

Proposal: Amending the GAIN AI Act to Incorporate KPIs to Track the AI Boom

While Sam Altman asserts that AI technology will drive a new wave of unprecedented economic growth, President Trump’s tariffs, as previously discussed, create potential uncertainty about the long-term impacts on the U.S. AI market and supply chain. Given this volatility, Congress must adopt mechanisms to monitor the domestic AI market. The GAIN AI Act offers one such approach; however, its effectiveness depends on measurable indicators of market performance, such as key performance indicators (KPIs). Accordingly, the Act should be amended to require the collection of KPIs to track the U.S.’s advancements in the AI sphere. KPIs are measurements used to assess a company’s long-term performance and progress on key business objectives. They are essential because they allow policymakers to evaluate whether incentives are producing the intended outcomes without imposing heavier forms of regulation. KPIs also create transparency for investors, developers, and agencies regarding how authorized U.S. AI companies purchasing advanced AI chips are growing within the market. 

This proposed amendment would integrate seamlessly into the Act’s existing framework, as the Act already codifies the definition between authorized versus unauthorized purchasers, including their application requirements to hold a license to purchase the AI chips. Additionally, KPI reporting can be incorporated into ongoing obligations of authorized purchasers, which could also supply the empirical backbone needed to evaluate whether the Act is truly stimulating domestic innovation. 

A similar structure exists in the CHIPS and Science Act of 2022, which codified its market-based approach by funding the Creating Helpful Incentives to Produce Semiconductors (CHIPS) for America International Technology Security and Innovation Fund. The Act assigns CHIPS and the Fund to provide “international information and communications technology security and semiconductor supply chain activities, including to support the development and adoption of secure and trusted telecommunication.” By codifying this responsibility, the assignees are obligated to track and foster technological innovation. By explicitly tying funding to measurable technological development, the CHIPS Act ensures accountability and strategic direction. 

Without a comparable KPI-focused provision in the GAIN AI Act, the Act risks distributing access to chips without the necessary mechanism to evaluate whether it is actually accelerating U.S. AI leadership, weakening its inertia as a newfound regulatory scheme. 

Proposal: A Statutory Safe Harbor for Investors to Ensure Capital Growth

To give investors greater confidence in backing their wagers while promoting the U.S.’s regulatory framework, Congress should also establish a statutory safe harbor that protects investors from secondary copyright liability arising from the conduct of AI developers or platforms they fund. An ideal safe harbor may have liability boundaries and limited liability for bad-faith actors. It could also force companies to implement ethical and technical guardrails to mitigate the production of infringing or deceptive outputs. Additionally, it may also force AI developers to publicly document their data sourcing, licensing, and model training protocols, though this is less likely to be practical. Overall, there needs to be a balance established between capital growth and accountability, and between innovation incentives and intellectual property integrity to stabilize the market for products like Sora 2. 

Conclusion

Sora 2 exemplifies the rapid innovation and market disruption enabled by AI, but it also exposes gaps in legal, regulatory, and investment frameworks. While U.S. policies, such as President Trump’s tariffs on specialized AI chips and the U.S. GAIN AI Act, support domestic development, uncertainty remains regarding copyright liability and market accountability. By amending the GAIN AI Act to require KPIs and establishing statutory safe harbor provisions, Congress can provide the empirical oversight and legal protections necessary to sustain AI growth and innovation while ensuring investor confidence, all of which is necessary to maintain the U.S.’s global competitiveness in AI.

*The views expressed in this article do not represent the views of Santa Clara University.