It’s not uncommon for technology to outpace laws, regulations and policies. In fact, this routinely occurs with every major technological shift, often causing policymakers to scramble. But in the case of generative artificial intelligence (AI), the speed of change seems greatly accelerated compared to the past. In only a few short months, numerous industries are embracing large language AI models like ChatGPT. In the process, many AI legal issues are being raised, along with concerns regarding AI legal protections. And it’s not only the U.S. that’s experiencing these hassles. Countries across the globe are trying to play catch-up when it comes to these new AI advances.
There are two major areas that deserve attention. The first, and perhaps the most pressing, involves how nations will oversee AI platforms as they permeate all aspects of society. Without adequate transparency and oversight, the potential for disinformation campaigns and unethical practices to disrupt societies is noteworthy. Secondly, it remains unclear where AI-generated content falls within copyright and intellectual property areas. Do AI legal protections exist for such content? If not, do new laws and regulations need to be created to govern these issues? These are the key issues that current comprise the tangled legal mess that AI content has generated.
“Responsible AI systems could bring enormous benefits, but only if we address their potential consequences and harms. For these systems to reach their full potential, companies and consumers need to be able to trust them.” – Alan Davidson, Director of U.S. National Telecommunications and Information Administration
National Approaches to AI Legal Issues
In the U.S., the Senate recently began discussing strategies regarding AI legal issues amidst its rapid rise in use. However, it seems they are lagging behind other countries in this regard. China has already developed policies that govern AI legal protections and are implementing them currently. Under the Cyberspace Administration of China (CAC), several procedures are now required among AI platforms. For one, all AI operators must submit an application subject to a safety review before offering their platform to the public. In addition, these same operators must reveal their AI algorithm and prove they are not engaging in unfair competition. Finally, the CAC is allowed to provide oversight of all content in order to ensure the accuracy of the information. As is evident, China is taking these AI legal protections seriously.
At the same time, Europe appears to be taking a similarly aggressive approach to AI legal issues. EU countries are proposing the AI Act, which imposes sweeping regulations over AI-generated content. The new laws would require the source identity of all AI-generated content while noting some content is unacceptable. The EU already has some of the strictest policies protecting individual privacy rights. It would seem that the AI Act would mirror previously passed legislation in this regard. As is evident for both the EU and China, there is concern that AI content could undermine national unity. It might also have the potential to fuel violence, discrimination, and terrorism. Given issues associated with AI content producing biased and false accounts, new policies are trying to prevent this from occurring.
As far as the U.S. is concerned, only recently has guidelines been suggested to address these same AI legal issues. Given that Google and Microsoft along with Open AI are leading the way in AI-generated content platforms, this is unfortunate. AI legal protections are needed, especially in light of the fact that its use in the country is already impressive. Not only are some news outlets using ChatGPT to a large extent, but so are some legal firms. As such, current policy proposals are suggesting that AI algorithms, intended audiences, and data sources be disclosed publicly. Likewise, detailed explanations about the AI content process and conclusion-drawing procedures have been considered. These discussions are in the early stages, however, in the U.S. with no immediate legislative action in sight.
“[China] is concerned about the potential for these large language models to generate politically sensitive content.” – Jeffrey Ding, Assistant Professor at George Washington University
AI Legal Protections of Content
In addition to national-level concerns about AI legal protections, there additional ones concerning content rights. Specifically, there are questions regarding whether or not AI-generated content can be subject to copyright laws. On the one hand, some have raised questions if a user of a AI platform might subsequently copyright its creations. Or, alternatively, does such content belong to parent companies like Open AI for ChatGPT? On the flip side, what if AI content caused harm as in the case of slander, defamation or falsehoods? Who is held responsible in these instances if AI generated the information? These are complex AI legal issues that need to be clarified as we enter this brave new world.
As it turns out, U.S. law states all AI-generated content is public domain material upon its creation. As such, the person using the AI platform as well as the parent company of the AI have no potential for copyright protections. Because AI content is not human-created directly, precedent suggests there is not such a capacity. However, in terms of publishing AI content, there may still be risks for those involved. The entity or person promoting AI-generated content might still be liable for harms that might occur. This is particularly true if the information is known to be false. Despite AI creating the materials, persons may lack legal protections if they knowingly propagate harmful content.
“The work must be the result of original and creative authorship by a human author. Absent human creative input, a work is not entitled to copyright protection.” – Margaret Esquenet, Partner with Finnegan, Henderson, Farabow, Garrett & Dunner, LLP.
A Legal Quagmire at Present
It seems pretty clear that the benefits that AI platforms might offer are tremendous. Large language models and other related AI systems could transform nearly every industry and do so quickly. It is precisely this potential that is raising many AI legal issues and causing nations to implement AI legal protections. But doing so will be difficult since overzealous government oversight could squash creativity and innovation. Certainly, some safeguards, accountability and transparency are needed. However, this must be balanced with technological progress that benefits societies throughout the world. Ideally, having time to sort these issues out gradually would be preferred. But this is simply not a luxury nations have currently.