By John Q. Hosedrinker 
To combat the misuse of AI-generated deepfakes, the Danish government plans to grant citizens property rights over their likeness and voice. This proposed legislation would allow individuals to request the removal of deepfakes featuring their image or voice from hosting platforms. The bill, which has cross-party support and is expected to pass this fall, aims to protect artists, public figures, and ordinary people from digital identity theft, addressing concerns highlighted by instances like AI-generated music mimicking popular artists. Further legislation is planned to potentially fine companies that fail to comply with takedown requests, reflecting Denmark’s commitment to both freedom of speech and individual rights in the age of generative AI.
Read the original article here
Denmark plans to thwart deepfakers by giving everyone copyright over their own features, and that’s definitely a headline that grabs your attention. The idea is to empower individuals by granting them ownership of their likeness, essentially giving them the right to control how their physical features are used and replicated, especially in the digital realm. It’s a bold move designed to combat the growing threat of deepfakes, which can be used for everything from harmless entertainment to malicious fraud and defamation.
The question immediately arises: Can this be done with our data too? It’s a valid concern. We’re talking about a fundamental shift in how we perceive and control our digital identities. The US, and other countries, should consider this approach. The proliferation of deepfakes on social media, particularly, is a growing problem.
There are obvious questions that need answering. What happens if people look alike? Are twins going to immediately be at odds with each other, potentially filing lawsuits over who has the right to their shared features? What if someone slightly alters someone’s features in a deepfake? Where do the lines get drawn?
Giving everyone copyright over their own features, it’s the core of this. Does this mean we didn’t have this control before? This could have huge implications. What about creating art that depicts someone? Will you be able to make a movie about someone? Imagine the legal challenges that could arise. The idea of being able to create an NFT (non-fungible token) of your own likeness also comes to mind.
There are some potential issues, for sure. How does this new law address international copyright? There’s the worry that people in other countries will simply ignore it. It’s also an issue of practicality. As one person put it, “you can’t outlaw math.” You can’t simply stop people from creating AI or using it. It might mean overloading the courts with a barrage of legal claims.
The practical implications of this are many. Can you take a photo in public, knowing that everyone in that shot has to give their permission? What about the background? And if that’s the case, how will this impact the film industry, for example, or photographers capturing moments in public spaces? The law’s impact will go way beyond the initial intention.
The focus on combating deepfakes, with their tendency towards malicious content, becomes critical. But, as one person pointed out, deepfakes are already used to create illegal and/or fraudulent content. Perhaps, it would be better to allow social media to become flooded with deepfakes and AI-generated content. Then, as the content becomes overwhelming, people might begin to rely more on actual news networks for their information.
Of course, some think the focus should be on regulating the AI creators, mandating that they own 100% of the content used by the AI for profit. Others have suggested the EU should enforce AI data collection to be opt-out by default instead of opt-in. These measures might be seen as a way to tackle the problem on a larger scale.
This also begs the question of what the legal repercussions would be if a deepfake actually sues the person whose likeness it mimics. This is the type of convoluted situation that might arise. And, of course, there is also the issue of parody. Protecting satire and humor is an important part of free speech, which means the law would need to carefully delineate what constitutes protected parody versus actionable defamation.
This could make copyright law very tricky. Consider voice actors, or people who do impersonations. These individuals might need to buy the rights before doing an impression. One can also see how the law could hinder biopics that aren’t authorized by the subject.
What about the potential for this to be weaponized? As mentioned, some feel the government is creating rules that will hinder the AI race. And, if Denmark implements this, server farms could be created in countries that see copyright as more of a suggestion. But, even if it is not used for malicious means, it’s clear that this law could be extremely difficult to enforce, given the interconnectedness of the digital world. And for some people, it feels like it just might be a stupid thing to do.