Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Fining Big Tech isn't working. Make them give away illegally trained LLMs as public domain

Dec 22, 2024 - theregister.com
The article discusses the ethical and legal challenges posed by the misuse of large language models (LLMs) like OpenAI's GPT, which are often trained using data scraped from the internet without consent or compensation. The author, Alexander Hanff, highlights the environmental impact of deleting these models under the "fruit of the poisonous tree" doctrine, which would render illegally obtained evidence inadmissible. Instead, Hanff suggests placing these models in the public domain if companies are found to have broken the law, thereby removing their economic advantage and ensuring compliance with legal obligations. This approach would require global cooperation and potentially new legislation to enforce effectively.

Hanff argues that current penalties are insufficient deterrents for tech giants, who often treat fines as a cost of doing business. He advocates for stronger regulatory measures to prevent companies from profiting from illegal data practices, emphasizing the need for global treaties and enhanced legal frameworks like the EU's Digital Markets Act. Hanff plans to lobby for legislative changes in Brussels to ensure that companies respect privacy and data protection laws, aiming to prevent the repetition of past regulatory failures and protect fundamental human rights.

Key takeaways:

```html
  • The author argues against deleting AI models trained on unlawfully obtained data due to the significant environmental impact of retraining them.
  • Proposes placing AI models into the public domain if they are found to have been trained unlawfully, as a deterrent to companies profiting from illegal data usage.
  • Highlights the need for global cooperation and stronger legal frameworks to enforce compliance with data protection laws and prevent unlawful AI training.
  • Emphasizes the importance of learning from past regulatory failures to ensure effective deterrents and consequences for companies that violate legal obligations.
```
View Full Article

Comments (0)

Be the first to comment!