What Seedance 2.0 Reveals About the Global AI Race

Chinese AI models are advancing rapidly — and U.S. policy choices will help determine whether America leads the next phase of AI development or cedes ground to competitors abroad.

Driving the news:

Seedance 2.0 is one of the latest reminders of how quickly advanced AI systems are emerging outside the United States. Released earlier this year, the Chinese video model drew attention for generating hyperrealistic clips, sparking backlash from U.S. lawmakers and Hollywood over concerns about copyright and the use of creative works in AI training.

• The response: Sens. Marsha Blackburn, R-Tenn., and Peter Welch, D-Vt., called on ByteDance to shut the model down, arguing that viral outputs reflected “the theft of American creative work.”

Why it matters:

Seedance is part of a broader acceleration in Chinese AI development. Companies such as ByteDance and Kuaishou are gaining ground in AI video, with models like Seedance 2.0 and Kling producing faster, sharper, and more realistic videos.

• The broader race: Recent advances from companies such as DeepSeek show that Chinese AI companies are also moving quickly on many fronts, not just video.

The stakes:

The United States still has major advantages, including world-class AI companies, leading universities, and a strong culture of innovation. But the global AI race is increasingly being shaped by speed, scale, and access to data.

Dive deeper:

Chinese firms are moving quickly, with significant state support and a national strategy focused on accelerating AI deployment. As part of that effort, China has launched an open-data campaign to make more public, academic, industry, and government-held datasets available for its AI companies, reflecting a growing recognition that access to data is a key advantage in the global AI race. The U.S. should not seek to replicate that model. America’s advantage comes from a different system — one rooted in the rule of law, open research, private-sector innovation, and responsible safeguards.

• The risk: If U.S. policymakers impose broad restrictions on lawful AI training, they risk weakening the very system that has helped America lead.

• The reality: If the U.S. slows AI development, it will not stop the global race. It will simply shift innovation, investment, and standard-setting power abroad.

Between the lines:

Concerns about misinformation, privacy, copyright, and job disruption should be taken seriously. Creators want their work respected, workers want to understand how AI will affect their jobs, and the public deserves safeguards against misuse.

• The trade-off: Sweeping licensing restrictions on lawful AI training would not address harms directly. Instead, they could make American AI systems less accurate, less useful, and less competitive while leaving foreign competitors free to move ahead.

Why data matters:

AI models depend on access to broad, diverse data to understand language, images, and real-world context. Limiting that access could make AI systems less capable and less representative of the people they are meant to serve.

• The market impact: Restrictive licensing mandates would favor the largest companies that can afford expensive data deals while making it harder for smaller organizations to compete.

• The expert view: At a Senate Commerce Committee hearing, Brad Smith, vice chair and president of Microsoft, argued that accessible public data “levels the playing field” for startups, academic institutions, and nonprofits.

The risk:

The countries building the most widely adopted AI systems will have an outsized influence over the standards and norms that shape the technology globally. Those systems will increasingly affect how information is created, how businesses operate, how governments deliver services, and how countries project influence.

• The smart approach: The U.S. can address legitimate AI harms without adopting blunt restrictions that weaken the broader innovation ecosystem. Policymakers should focus on targeted safeguards that protect creators, consumers, and workers while preserving the lawful pathways that allow American AI development to continue.

Bottom line:

The global AI race is already underway. U.S. policymakers should protect creators and consumers without cutting off lawful pathways for AI development. That means preserving broad access to data, supporting open and competitive AI ecosystems, investing in research and infrastructure, creating clear national rules instead of a fragmented state-by-state patchwork, and ensuring startups, researchers, and universities can participate in the next wave of AI.

America can lead AI development with democratic values, but only if we keep building.