OpenAI wants the US government to establish ‘fair use' protections for AI training

According to a new report from CNBC, OpenAI's been busy stirring up the conversation about how the US government handles AI. The company recently submitted a proposal asking the U.S. government to codify the idea of "fair use" when it comes to training AI models.

It's worth noting that OpenAI has been walking a tightrope lately. The company is in a heated legal battle over copyright issues, having been sued by several major news outlets and creators for using copyrighted content to train models like GPT-4.

OpenAI's pitch is pretty clear about wanting to see fewer heavy-handed rules and more room to innovate. It argues that America's success in AI is partly thanks to the fair use doctrine that has allowed companies to learn from a wide variety of content. In fact, the company wrote,

America has so many AI startups, attracts so much investment, and has made so many research breakthroughs largely because the fair use doctrine promotes AI development.

This context is especially interesting considering the recent shuffle in executive orders. After President Trump scrapped the first AI executive order that was put in place during President Biden's term, OpenAI quickly moved to make sure it had a say in what comes next. The company wants to be part of shaping the new "AI Action Plan" and sees the government working hand in hand with both tech giants and startups.

Here are some of the key points the ChatGPT creator is pushing for:

• The government should create a partnership with private companies on a voluntary basis rather than enforcing overly strict state laws.
• There should be a faster, criteria-based path for AI tools to get federal security approvals. This could potentially cut down the wait time by about 12 months.
• OpenAI also recommends an "export control strategy" that would promote the global adoption of American AI systems.

Alongside these proposals, it is also raising concerns about potential risks, particularly when it comes to Chinese AI competitors like DeepSeek. The company warns that while America is still ahead in the AI race, the gap is narrowing fast because rivals might be cutting corners in ways that could compromise user privacy and security.