The race between China and the United States in artificial intelligence has entered a new and unsettling phase. It's not just semiconductors or chatbots anymore, but something much more powerful: high-quality video content and the ability to fabricate reality itself, a Newsweek analysis says.
In recent days, ByteDance's newest AI model, Seedance 2.0, has sparked strong reactions on the internet. Users generated a hyper-realistic clip of 'Tom Cruise' and 'Brad Pitt' fighting on a rooftop over 'Jeffrey Epstein', starting with just a simple text prompt. The sequence looked cinematic with fluid movements and synchronized dialogue. There were no actors, cameras or film crew – just code.
The implications go far beyond the Hollywood industry. Seedance 2.0 marks the entry into the era of what pundits call “high quality slopaganda” – synthetic content cheap to produce, emotionally charged and realistic enough to pass a cursory inspection. Unlike the clumsy deepfakes of a few years ago, the new generation of AI videos is extremely convincing.
A new political weapon
In politics, Seedance 2.0 isn't just a technological leap in AI capabilities, it marks the emergence of something that can be skillfully used as a new political weapon.
The moment introduces a volatile dynamic into the landscape of American politics. For President Donald Trump — a figure who has mastered the viral spectacle and the mechanisms of the attention economy — this technology represents both an opportunity and a risk.
His political rise has been fueled by the ability to dominate news cycles and digital platforms. But in a world where anyone can fabricate convincing images of themselves saying or doing almost anything, control becomes fragile. A fake clip that portrays it as confusing, extremist or contradictory could spread to millions of users before fact-checkers intervene.
Even after removing the fakes, the effects may persist. Repetition leaves cognitive traces. The danger is not just persuasion, but the erosion of trust: if voters begin to question the authenticity of any political video material, trust in public communication can collapse. In such a climate, even the most seasoned media figures run the risk of being overwhelmed by a continuous stream of persuasive noise.
The mid-term elections, a digital minefield
The stakes are all the higher in view of the upcoming US midterm elections. These elections are often decided on the margins, in a few key districts where turnout is crucial. A fabricated video targeting a Senate or House candidate could circulate for hours or days before platforms react.
AI-generated phone calls could perfectly mimic a candidate's voice. “Leaked” records could appear even on the eve of the vote. Even if they are later dismantled, the initial impact could discourage participation or fuel accusations of illegitimacy. The objective would not necessarily be to change beliefs, but to inject confusion at critical moments.
Foreign adversaries, alert to opportunity
If such tools are widely available in China, they are unlikely to remain there. Advanced generative systems are spreading rapidly – through open-source imitation, commercial partnerships or parallel development. And if one power can build them, others can turn them into weapons.
Russia, Iran, and North Korea have already demonstrated over the past decade that they are willing to exploit online ecosystems to sow division in the United States and its allies. From election interference to escalating culture wars, these states have shown that they don't need to win over a majority of voters to succeed. It is enough to deepen mistrust and widen fissures in already polarized societies.
High-fidelity AI video tools would make these campaigns more scalable, more compelling, and much harder to dismantle. The goal would not be to promote a single dominant narrative, but to overload the information system with so much contradictory and emotional content that citizens no longer trust anything.
Hollywood is sounding the alarm
The impact of technology is not limited to politics. Seedance 2.0 has sparked strong reactions in the entertainment industry as well. Users recreated famous characters and actors without consent, including Disney-style figures and characters from the Marvel and Star Wars universes. The resulting cinematic-looking clips immediately raised copyright issues.
Studios and industry organizations argue that such designs appear to be based on copyrighted material and the use of recognizable images, which could constitute widespread infringements of intellectual property. Artists' unions warn that digital replication of faces and voices threatens actors' livelihoods.
ByteDance has announced it is strengthening its safeguards, but the controversy highlights growing global tensions over AI, ownership and control of creative content.
When reality becomes contested
What sets this moment apart from previous waves of digital manipulation is its scale and sophistication. Artificial intelligence doesn't just alter images anymore – it creates entire scenes with believable dialogue and genuine emotion. The strategic advantage lies not necessarily in winning a single narrative battle, but in flooding the information space until authenticity becomes impossible to verify.
If voters can no longer reliably distinguish between real images and synthetic fictions, democratic debate becomes unstable. Common facts erode. Public trust is eroding.
In this sense, China's progress in AI is not just a major technological leap, but a signal of a future where influence can be automated, doubt becomes constant, and reality itself becomes a contested terrain.
