Google has released a policy proposal that calls for balanced copyright rules, aiming to enable AI developers to use publicly available data without lengthy negotiations. The company argues that access to public data is crucial for improving generative AI systems. However, content owners are taking action against Google, with the tech giant facing several lawsuits.
The issue of copyrighted data usage in AI development is a complex one. Google claims that its models require vast amounts of publicly available data to function effectively, but this also raises concerns about data rights and ownership. The company’s proposal aims to address these concerns by introducing flexible regulations that balance the needs of rightsholders with those of AI developers.
Google also emphasizes the importance of modernizing the US energy infrastructure to support its growing data center power demand. The company projects a 40-gigawatt increase in global data center power demand from 2024 to 2026, citing current infrastructure limitations as a major obstacle.
To promote the development and adoption of AI systems, Google advocates for government-led initiatives, such as adopting multi-vendor approaches, releasing public datasets, and funding early-stage research. The company hopes that by leading by example, governments can create an environment that fosters innovation and cooperation between industry and academia.
Critics argue that the US’s current laws do not provide adequate protections for ordinary citizens, leaving them vulnerable to exploitation by powerful corporations. If the government fails to address these concerns, Google’s proposal may fall short in addressing the fundamental issues at stake.
Source: https://arstechnica.com/google/2025/03/google-agrees-with-openai-that-copyright-has-no-place-in-ai-development