How the world is (or isn’t) trying to rein in the wild west of artificial intelligence

Good morning. I know what you’re thinking: AI again?!

Though it can feel tiresome, there is a reason behind all of the fervour this technology creates. AI, a term that encompasses a number of emergent tools inspired by human intelligence, has the potential to redefine and transform almost all aspects of life. From personalised medicine to increased powers in policing, AI has become increasingly integrated in the way society functions, bringing with it a slew of malign side effects that have only recently become apparent.

Scientists have been working on AI for decades, and that is where the story starts in the new Guardian podcast Black Box: at its genesis 50 years ago. The seven-episode series illuminates what in reality is an incredibly complex and at times impenetrable tech story through human experiences. The second episode, in which host Michael Safi and producers Alex Atack and Joshua Kelly search for the creators of deepfake porn app ClothOff, is an amazing listen. You can subscribe to the podcast here.

For today’s newsletter, I spoke with Joshua Kelly and Prof Lilian Edwards, a leading academic in the field of internet law, about the way artificial intelligence is shaping society and how it could be regulated in the future. That’s right after the headlines.

Five big stories

In depth: ‘There’s no accountability in this space’

Initially, the team behind Black Box cast a wide net into the world of AI in order to draw out some themes and areas of particular interest. What they ended up finding was “the early days of a collision between a new technology and the existing shape of society”, Joshua says.

A case in Spain

One of the biggest challenges that has emerged from this technology is AI-driven synthetic media, more commonly referred to as deepfake technology. Deepfakes have become more sophisticated over the past decade, with AI making it easier than ever to produce convincing fake images and videos.

In Black Box, the team went to Spain to report on how this technology has caused mayhem in a small rural town, after parents realised that teenage boys were creating nude and explicit images of their female classmates on an AI app called ClothOff.

“We imagined that a single person, like maybe a sad guy in his basement, was making a little bit of money off the app but we realised it was much more sophisticated than that,” Joshua says. Those who run ClothOff likely run other similar apps, and over the course of their reporting Joshua noticed an increasing number of them.

The supply is meeting a growing demand: ClothOff alone had more than 700,000 subscribers on its Telegram channel and 12m views in the last three months. Joshua adds that ClothOff was brazen about its rapid growth: “They told us their value had doubled since we first started talking to them in autumn.”

Back in Spain the local authorities in the rural town didn’t know how to deal with the problem: those using the app were all minors, which complicated criminal prosecution, and those who created the app were operating from a foreign country. “We wondered if it was true that there’s no accountability in this space,” Joshua says.

Who is regulating AI?

As the consequences of artificial intelligence grow clearer and more existential with every passing day, there has been a growing concerted effort to put up safeguards. Different countries are taking different approaches. Here are a few of the biggest players:

EU: The European Union has been the leading player on the global stage when it comes to AI regulation. The bloc recently passed the Artificial Intelligence Act (AIA), the world’s first such comprehensive law, which comes into effect in 2025. It includes new rules on the use of facial recognition technology in public places by the police, large generative models like ChatGPT and many other aspects of AI. Even though it is now “the gold standard model in the world”, Lilian Edwards says that the AIA is still fairly limited.

US: The AIA has “spooked” the US, Edwards says, because it forces American companies that are trying to sell in the EU, the second largest market in the world, to meet its standards. “It’s effectively a kind of legal land grab by the EU,” Edwards adds.

This, coupled with a president who has made it a priority to rein in big tech companies, has changed the tenor of the conversation in the US, which has historically been extremely anti-regulation when it comes to tech. True to his word, Biden issued an ambitious executive order that purportedly took on the safe deployment of AI in October and has given more power to NIST, the National Institute of Standards and Technology.

UK: In short, the UK’s stance is to do nothing. The government wants to be the “Singapore of tech”, Edwards says, and so has taken an anti-regulatory, “pro-innovation” stance with the hopes that Silicon Valley giants will come flocking to Britain, though, so far, not much has happened.

China: China is already one of the world leaders in AI research and has been trying to tackle head on what that could mean for its society. It has spearheaded a vertical legislation approach, meaning that instead of pushing through one very broad law like the AIA, the Chinese government has been trying to regulate individual aspects of AI.

Are we all doomed?

The evolving nature of AI will be challenging to regulate within institutions that are by nature slow moving, but Edwards believes that the difficulty can be overblown. Not only are politicians and regulators creating new legislation in many countries, there are also existing legal frameworks that can and do protect people from AI.

“Copyright and privacy laws and the rules of product liability, consumer protection and consumer safety, all already exist. When you apply this huge amount of existing law to novel technologies, sometimes it works and sometimes you get snagging problems that you have to amend and change,” Edwards says, but that does not mean that the situation is a total free for all.

The longer it takes though, the more harrowing the consequences could be. In the small Spanish town at least 20 girls between the ages of 11 and 17 had deepfake nude images created of them, and the effects were profound and devastating. “Some of the girls felt they were unable to leave their homes, their academic performances were dropping, they were suffering from anxiety attacks,” Joshua says. And this was just one group of children, in one place, who used one app. “It wouldn’t be surprising if, this time next year, there were 10 times as many [of these apps] that are even more powerful than what we’re seeing now.”

For more on AI and how tech is changing the world around us, click here to get Alex Hern’s TechScape newsletter every Tuesday

What else we’ve been reading

• There’s a mountain of Oscars coverage to sate your hunger this morning: Peter Bradshaw’s view of “an Oscars night with a British accent – and no real upsets”, a round-up of the most entertaining viral moments, pictures from the ceremony and the best red carpet looks (including an Ariana Grande outfit memorably described as “bring your duvet to work day”), and a full list of winners. Archie

• It was years into his career as a teacher until Michael Donkor felt able to come out to his pupils. He wrote a superb piece for Saturday magazine about what stopped him from doing so, and the impact when he finally did. Archie

• “I was cancelled”: As a chart-topping singer in the ultra-controlled pop landscape of the 00s, Beverley Knight refused to be sexy, spoke out about Aids and was staunchly pro-Labour … and now she wants the Tories out! Nazia Parveen, acting deputy editor, newsletters

• In recent months, marches and protests over Gaza and the climate crisis have regularly been depicted as illegitimate challenges to democracy. Andy Beckett writes a necessary and urgent piece arguing that, in fact, “when a parliament fails to speak for enough voters, politics takes other forms”. Archie

• She was a former debutante who turned to terrorism: Sean O’Hagan takes a fascinating deep dive into the life of Rose Dugdale, an English heiress who became an IRA bomber and art thief whose life story is due to be dramatised in a new film. Nazia

Sport

Football | An Alexis Mac Allister penalty cancelled out John Stones’ first half opener to secure Liverpool a 1-1 draw in a thrilling top-of-the-table clash against Manchester City. Meanwhile, Tottenham beat Aston Villa 4-0 to close the gap in the race for a top-four finish, West Ham and Burnley drew 2-2, and Brighton beat Nottingham Forest 1-0.

Rugby | Wales gave up a 24-20 lead to be overpowered by France 24-45, with three tries from the visitors in the final quarter sweeping the hosts away. The result means that Wales will enter their final Six Nations match against Italy at risk of finishing with the wooden spoon.

Cricket | Brendon McCullum has admitted that his England side must be “a bit smarter” with their comments in public after crumbling to a 4-1 series defeat against India even as players boasted of their confidence. In reference to remarks like Ben Duckett’s suggestion that “the more, the better” for England facing a run chase, he added: “It is not arrogance, just confidence in the group.”

The front pages

The Guardian splashes on “Tory ex-home secretaries warn against politicising extremism”. “Blasphemy demos ‘are growing in radicalism’” reports the Times, which for its picture lead has the Princess of Wales and her children – controversial because it’s apparently been manipulated, which the paper highlights with a yellow circle. That’s the top story in the Daily Telegraph: “Photo from palace was doctored, say agencies” – that’s in reference to the wire services that recalled the picture after realising something was up.

Others seem none the wiser. “I’m smiles better” – the photo is on the front of the Metro, but the allegedly doctored bit is cropped out. “Thank you” says the Daily Mirror, conveying the princess’s sentiments, while the Daily Express fawns at this “charming photo that proves she’s on the mend” (its news splash is “We’ll get tougher with work-shy to cut taxes, says PM”). “Town halls’ woke spree” – that’s the Daily Mail, which is awake to the manipulated royal photo and asks of the photo op: “has it backfired”? “Biden struggles to win voter credit as Americans feel happier on economy” says the Financial Times. “Faulty IT at Home Office hits effort to tackle illegal immigration”, reports the i.

Today in Focus

Cartoon of the day | Edith Pritchett

Sign up for Inside Saturday to see more of Edith Pritchett’s cartoons, the best Saturday magazine content and an exclusive look behind the scenes

The Upside

A bit of good news to remind you that the world’s not all bad

Is there anybody out there? Well, there just might be. Astronomers have detected a distant planet – described as a “water world with a boiling ocean” – in deep space, a significant discovery by Nasa’s James Webb telescope that advances the search for habitable conditions beyond Earth.

The observations revealed water vapour and chemical signatures of methane and carbon dioxide in the atmosphere of the exoplanet, which is twice Earth’s radius and about 70 light years away. It’s a chemical mix consistent with a water world where the ocean would span the entire surface, reckon researchers from Cambridge university.

Sign up here for a weekly roundup of The Upside, sent to you every Sunday

Bored at work?

And finally, the Guardian’s puzzles are here to keep you entertained throughout the day. Until tomorrow.

• Quick crossword

• Cryptic crossword

• Wordiply