Summary of the Core Content
This news article discusses the impact of AI cloning and dubbing technology on the livelihoods of ordinary voice actors. Individuals like Xiaoyu, who earn around 10,000 yuan per month from their work, face a significant drop in income and no legal avenues for redress due to unauthorized use of their voices by AI. The cost of AI dubbing is extremely low, allowing partners to negotiate lower prices. The proliferation of cloned voices has led to real voices being misidentified as AI by algorithms, preventing them from entering the content distribution streams. Legal frameworks struggle to address minor instances of copyright infringement across the entire internet. Moreover, this phenomenon has given rise to social issues such as AI voice-related fraud. Seventy percent of small and medium-sized practitioners in the industry have seen their incomes halved, and even top-tier actors are affected. Platforms and clients prefer AI-generated voices, leaving real voice actors in a predicament where it's difficult to distinguish between genuine and fake content, with no one to uphold justice.
How Does AI Dubbing Take Away Voice Actors' Jobs?
Voice actors like Xiaoyu are at the bottom of the industry chain; they receive only 1.5 yuan per minute from the platform after the client has paid 30 yuan. With the advent of AI, the cost of deploying these models is virtually zero, saving considerable amounts for accounts that produce hundreds of pieces of content each month. Partners directly negotiate prices using the AI-generated versions of Xiaoyu's voice. Although real-person dubbing is more expensive, AI offers an unlimited supply of high-quality voices at low costs. Industry research shows that after voice theft, the income of 70% of small and medium-sized actors has been reduced by more than half, and even top-tier actors have seen a 30% drop. What's left for these actors is often to serve as training data for AI models.
Why Is It So Difficult to Protect Rights?
In 2024, the Beijing Internet Court ruled on the first case of AI voice infringement in China, with the voice actor Yin winning because he was able to identify the specific company responsible. However, Xiaoyu's situation is different; his voice is spread across hundreds of media accounts, making it impossible to trace the source of the theft, and each instance amounts to a small amount of infringement. The law can handle cases where one person sues a single company but cannot address situations where an individual's work is stolen by the entire internet. What's more frustrating is that when large organizations use Xiaoyu's AI-generated voice, people may assume it's always AI anyway, making it seem as if Xiaoyu is exaggerating his claims.
The Identity Crisis of Real People Being Hindered by Their Own Replicas
The more cloned versions of Xiaoyu's voice are circulated, the more likely algorithms are to mistake his real voice for an AI one, preventing it from being included in content distribution streams. This is similar to how the Six-Eared Monkey transforms into Sun Wukong—no magic mirror can distinguish between the two. Platforms don't have the resources or risk of making incorrect judgments, so they simply categorize all such voices as AI to save time and effort. Partners also prefer the AI version of Xiaoyu because real people can get tired, refuse to work on certain projects, or lose motivation, while AI remains available 24/7.
The Limits of AI Dubbing: Where Can It Be Used?
Not all scenarios are suitable for AI dubbing. For example, documentary director Helen may use AI for internal demos (with limited budget and non-commercial purposes), but not for official documentaries, which require the "imperfections" inherent in human performance—pauses, breaths, and hesitation. These elements add authenticity, and nothing can replace them, even with advanced AI technology. However, these seemingly minor uses can accumulate and pave the way for AI to replace real voices.
The Threat of AI Voices in Fraud
The dangers of AI-generated voices extend beyond disrupting the industry. In 2023, someone in Fuzhou used an AI voice to impersonate a businessman and defrauded 4.3 million yuan; in 2024, a Hong Kong employee transferred 200 million Hong Kong dollars to a fake CFO created by AI. Xiaoyu worries that his cloned voice could be used to scam his parents. These cases highlight how AI voices can become tools for fraud, posing threats to both individuals and businesses.
In conclusion, Xiaoyu's situation resembles the journey of Tang Seng in "Journey to the West"—without a definitive authority to determine what is genuine or fake, real voice actors are left without recourse. Even if they win all legal battles, people may still remember the AI-generated version of their voices years later. This is the most frustrating aspect of the situation.