Denmark eyes new law to protect citizens from AI deepfakes

COPENHAGEN, Denmark -- In 2021, Danish video crippled live-streamer Marie Watson received an representation of herself from an chartless Instagram account.

She instantly recognized the vacation drawback from her Instagram account, but thing was different: Her covering had been digitally removed to marque her look naked. It was a deepfake.

“It overwhelmed maine truthful much,” Watson recalled. “I conscionable started bursting retired successful tears, due to the fact that suddenly, I was determination naked.”

In the 4 years since her experience, deepfakes — highly realistic artificial intelligence-generated images, videos oregon audio of existent radical oregon events — person go not lone easier to marque worldwide but besides look oregon dependable exponentially much realistic. That's acknowledgment to technological advances and the proliferation of generative AI tools, including video procreation tools from OpenAI and Google.

These tools springiness millions of users the quality to easy spit retired content, including for nefarious purposes that scope from depicting celebrities Taylor Swift and Katy Perry to disrupting elections and humiliating teens and women.

In response, Denmark is seeking to support mean Danes, arsenic good arsenic performers and artists who mightiness person their quality oregon dependable imitated and shared without their permission. A measure that's expected to walk aboriginal adjacent year would alteration copyright instrumentality by imposing a prohibition connected the sharing of deepfakes to support citizens’ idiosyncratic characteristics — specified arsenic their quality oregon dependable — from being imitated and shared online without their consent.

If enacted, Danish citizens would get the copyright implicit their ain likeness. In theory, they past would beryllium capable to request that online platforms instrumentality down contented shared without their permission. The instrumentality would inactive let for parodies and satire, though it’s unclear however that volition beryllium determined.

Experts and officials accidental the Danish authorities would beryllium among the astir extended steps yet taken by a authorities to combat misinformation done deepfakes.

Henry Ajder, laminitis of consulting steadfast Latent Space Advisory and a starring adept successful generative AI, said that helium applauds the Danish authorities for recognizing that the instrumentality needs to change.

“Because close now, erstwhile radical accidental ‘what tin I bash to support myself from being deepfaked?’ the reply I person to springiness astir of the clip is: ‘There isn’t a immense magnitude you tin do,’" helium said, ”without maine fundamentally saying, ‘scrub yourself from the net entirely.’ Which isn’t truly possible."

He added: "We can’t conscionable unreal that this is concern arsenic accustomed for however we deliberation astir those cardinal parts of our individuality and our dignity."

U.S. President Donald Trump signed bipartisan authorities successful May that makes it illegal to knowingly people oregon endanger to people intimate images without a person’s consent, including deepfakes. Last year, South Korea rolled retired measures to curb deepfake porn, including harsher punishment and stepped up regulations for societal media platforms.

Danish Culture Minister Jakob Engel-Schmidt said that the measure has wide enactment from lawmakers successful Copenhagen, due to the fact that specified integer manipulations tin disturbance doubts astir world and dispersed misinformation.

“If you’re capable to deepfake a person without her oregon him being capable to person that merchandise taken down, that volition undermine our democracy,” helium told reporters during an AI and copyright league successful September.

The instrumentality would use lone successful Denmark, and is improbable to impact fines oregon imprisonment for societal media users. But large tech platforms that neglect to region deepfakes could look terrible fines, Engel-Schmidt said.

Ajder said Google-owned YouTube, for example, has a “very, precise bully strategy for getting the equilibrium betwixt copyright extortion and state of creativity."

The platform's efforts suggest that it recognizes “the standard of the situation that is already present and however overmuch deeper it’s going to become,” helium added.

Twitch, TikTok and Meta, which owns Facebook and Instagram, didn't respond to requests for comment.

Engel-Schmidt said that Denmark, the existent holder of the European Union's rotating presidency, had received involvement successful its projected authorities from respective different EU members, including France and Ireland.

Intellectual spot lawyer Jakob Plesner Mathiasen said that the authorities shows the wide request to combat the online information that's present infused into each facet of Danish life.

“I deliberation it decidedly goes to accidental that the ministry wouldn’t marque this bill, if determination hadn’t been immoderate juncture for it,” helium said. “We're seeing it with fake news, with authorities elections. We are seeing it with pornography, and we’re besides seeing it besides with celebrated radical and besides mundane radical — similar you and me.”

The Danish Rights Alliance, which protects the rights of originative industries connected the internet, supports the bill, due to the fact that its manager says that existent copyright instrumentality doesn't spell acold enough.

Danish dependable histrion David Bateson, for example, was astatine a nonaccomplishment erstwhile AI dependable clones were shared by thousands of users online. Bateson voiced a quality successful the fashionable “Hitman” video game, arsenic good arsenic Danish toymaker Lego's English advertisements.

“When we reported this to the online platforms, they accidental ‘OK, but which regularisation are you referring to?’” said Maria Fredenslund, an lawyer and the alliance’s director. “We couldn’t constituent to an nonstop regularisation successful Denmark.”

Watson had heard astir chap influencers who recovered digitally-altered images of themselves online, but ne'er thought it mightiness hap to her.

Delving into a acheronian broadside of the web wherever faceless users merchantability and stock deepfake imagery — often of women — she said she was shocked however casual it was to make specified pictures utilizing readily disposable online tools.

“You could virtually conscionable hunt ‘deepfake generator’ connected Google oregon ‘how to marque a deepfake,’ and each these websites and generators would popular up,” the 28-year-old Watson said.

She is gladsome her authorities is taking action, but she isn’t hopeful. She believes much unit indispensable beryllium applied to societal media platforms.

“It shouldn’t beryllium a happening that you tin upload these types of pictures,” she said. “When it’s online, you’re done. You can’t bash anything, it’s retired of your control.”

___

Stefanie Dazio successful Berlin, Kelvin Chan successful London, and Barbara Ortutay successful San Francisco, contributed to this report.