Google has released a new policy proposal advocating for relaxed AI copyright restrictions and more flexible export controls. This move follows OpenAI’s similar recommendations and comes in response to the Trump administration’s call for a national AI Action Plan.
Google Advocates for Weak Copyright Restrictions in AI Training
In its proposal, Google calls for a regulatory framework that prioritizes AI innovation while minimizing restrictions. The company stresses that current AI policymaking has focused too much on risks, often at the expense of scientific progress and national competitiveness.
One of Google’s most debated recommendations involves using copyrighted content for AI training. Google asserts that fair use and text-and-data mining exceptions are essential for advancing AI research and model development. The company argues that training on publicly available, copyrighted data should be legally permitted without major restrictions.
“These exceptions enable AI models to use publicly accessible, copyrighted material for training without significantly harming rights holders,” Google states. The company warns that enforcing strict copyright laws could lead to unpredictable and lengthy negotiations that slow AI advancements.
However, Google’s stance has sparked legal challenges. Multiple lawsuits accuse the company of training its AI on copyrighted data without permission or compensation. U.S. courts have yet to decide whether AI developers are shielded under fair use laws.
Google Challenges AI Export Restrictions
Google also opposes certain export controls enacted under the Biden administration, claiming they could weaken U.S. economic competitiveness. The company argues that current regulations place an undue burden on American cloud service providers. This position contrasts with Microsoft’s, which recently stated that it can fully comply with existing export rules.
These controls primarily restrict the sale of advanced AI chips to specific countries. However, they do allow trusted businesses to access large-scale AI chip clusters, offering some flexibility.
Pushing for Federal AI Legislation and Increased R&D Investment
The company also urges the U.S. government to pass comprehensive federal AI regulations to address inconsistencies caused by the patchwork of state laws. According to tracking data, over 780 AI-related bills are currently pending across the U.S.
Google advocates for sustained investment in domestic AI research. It recommends that the federal government release datasets for AI training and allocate funds to early-stage AI projects. The company believes these steps will help ensure AI models and computing resources remain widely available to researchers and institutions.
Opposition to AI Liability and Transparency Requirements
Google raises concerns about AI liability laws, arguing that model developers should not be held responsible for how their AI is used. The company states that, in many cases, developers have little control over how third parties deploy their models. Instead, Google believes that responsibility should fall on those who integrate and use AI in real-world applications.
The company has historically opposed AI accountability measures, including California’s SB 1047, which sought to outline clear AI safety precautions. Google also pushes back against AI transparency requirements, particularly those introduced in the EU. The company argues that disclosing too much information could expose trade secrets, aid competitors, and even create security risks.
Growing Global Pressure for AI Regulation
Despite Google’s objections, many governments are moving toward stricter AI regulations. California’s AB 2013 mandates that AI companies publish summaries of the datasets they use, while the EU’s AI Act will require developers to provide detailed documentation about model capabilities and risks.
As global AI regulation continues to evolve, Google’s latest proposal highlights the ongoing debate between fostering innovation and implementing responsible oversight.
For more AI news, click here.
