Big Tech is Stealing Your Face

Big Tech is Stealing Your Face Generative AI models are increasingly trained on images and voice data scraped from the public internet, including social media and even private user information, without explicit consent. This raises significant issues regarding the scope, permanence, and autonomy of personal data, as once absorbed into AI systems, it becomes virtually impossible to track or remove. While some legal measures are being introduced to grant individuals rights over their digital replicas, current protections remain fragmented.

• Tech companies are using images and voice data from public and private online sources to train generative AI models.
• The ‘Scope Problem’ means you cannot predict future uses of your image across evolving technologies.
• The ‘Temporality Problem’ makes uploaded images permanent within AI systems, even if deleted from original sources.
• The ‘Autonomy Trap’ suggests AI trained on personal data can narrow individual options and representation.
• Companies like Scale AI and Clearview AI have been implicated in large-scale data scraping without consent.
• A ByteDance feature demonstrated AI generating a voice from a single facial photo, highlighting integrated media capabilities.
• Legal efforts include Matthew McConaughey’s trademarks, proposed Danish copyright law, the US TAKE IT DOWN Act, and the NO FAKES Act.
• The US Copyright Office has called for a federal ‘digital replica’ law to address AI-generated content.
• Current legal protections are a patchwork, leaving individuals vulnerable to unauthorized use of their likeness and voice. Continue reading https://gettingrealaboutai.substack.com/p/big-tech-is-stealing-your-face