More stories

  • in

    Tech firms sign ‘reasonable precautions’ to stop AI-generated election chaos

    Major technology companies signed a pact Friday to voluntarily adopt “reasonable precautions” to prevent artificial intelligence tools from being used to disrupt democratic elections around the world.Executives from Adobe, Amazon, Google, IBM, Meta, Microsoft, OpenAI and TikTok gathered at the Munich Security Conference to announce a new framework for how they respond to AI-generated deepfakes that deliberately trick voters. Twelve other companies – including Elon Musk’s X – are also signing on to the accord.“Everybody recognizes that no one tech company, no one government, no one civil society organization is able to deal with the advent of this technology and its possible nefarious use on their own,” said Nick Clegg, president of global affairs for Meta, the parent company of Facebook and Instagram, in an interview ahead of the summit.The accord is largely symbolic, but targets increasingly realistic AI-generated images, audio and video “that deceptively fake or alter the appearance, voice, or actions of political candidates, election officials, and other key stakeholders in a democratic election, or that provide false information to voters about when, where, and how they can lawfully vote”.The companies aren’t committing to ban or remove deepfakes. Instead, the accord outlines methods they will use to try to detect and label deceptive AI content when it is created or distributed on their platforms. It notes the companies will share best practices with each other and provide “swift and proportionate responses” when that content starts to spread.The vagueness of the commitments and lack of any binding requirements likely helped win over a diverse swath of companies, but disappointed advocates were looking for stronger assurances.“The language isn’t quite as strong as one might have expected,” said Rachel Orey, senior associate director of the Elections Project at the Bipartisan Policy Center. “I think we should give credit where credit is due, and acknowledge that the companies do have a vested interest in their tools not being used to undermine free and fair elections. That said, it is voluntary, and we’ll be keeping an eye on whether they follow through.”Clegg said each company “quite rightly has its own set of content policies”.“This is not attempting to try to impose a straitjacket on everybody,” he said. “And in any event, no one in the industry thinks that you can deal with a whole new technological paradigm by sweeping things under the rug and trying to play Whac-a-Mole and finding everything that you think may mislead someone.”Several political leaders from Europe and the US also joined Friday’s announcement. Vera Jourová, the European Commission vice-president, said while such an agreement can’t be comprehensive, “it contains very impactful and positive elements”. She also urged fellow politicians to take responsibility to not use AI tools deceptively and warned that AI-fueled disinformation could bring about “the end of democracy, not only in the EU member states”.The agreement at the German city’s annual security meeting comes as more than 50 countries are due to hold national elections in 2024. Bangladesh, Taiwan, Pakistan and most recently Indonesia have already done so.Attempts at AI-generated election interference have already begun, such as when AI robocalls that mimicked the US president Joe Biden’s voice tried to discourage people from voting in New Hampshire’s primary election last month.Just days before Slovakia’s elections in November, AI-generated audio recordings impersonated a candidate discussing plans to raise beer prices and rig the election. Fact-checkers scrambled to identify them as false as they spread across social media.Politicians also have experimented with the technology, from using AI chatbots to communicate with voters to adding AI-generated images to ads.The accord calls on platforms to “pay attention to context and in particular to safeguarding educational, documentary, artistic, satirical, and political expression”.It said the companies will focus on transparency to users about their policies and work to educate the public about how they can avoid falling for AI fakes.Most companies have previously said they’re putting safeguards on their own generative AI tools that can manipulate images and sound, while also working to identify and label AI-generated content so that social media users know if what they’re seeing is real. But most of those proposed solutions haven’t yet rolled out and the companies have faced pressure to do more.That pressure is heightened in the US, where Congress has yet to pass laws regulating AI in politics, leaving companies to largely govern themselves.The Federal Communications Commission recently confirmed AI-generated audio clips in robocalls are against the law, but that doesn’t cover audio deepfakes when they circulate on social media or in campaign advertisements.Many social media companies already have policies in place to deter deceptive posts about electoral processes – AI-generated or not. Meta says it removes misinformation about “the dates, locations, times, and methods for voting, voter registration, or census participation” as well as other false posts meant to interfere with someone’s civic participation.Jeff Allen, co-founder of the Integrity Institute and a former Facebook data scientist, said the accord seems like a “positive step” but he’d still like to see social media companies taking other actions to combat misinformation, such as building content recommendation systems that don’t prioritize engagement above all else.Lisa Gilbert, executive vice-president of the advocacy group Public Citizen, argued Friday that the accord is “not enough” and AI companies should “hold back technology” such as hyper-realistic text-to-video generators “until there are substantial and adequate safeguards in place to help us avert many potential problems”.In addition to the companies that helped broker Friday’s agreement, other signatories include chatbot developers Anthropic and Inflection AI; voice-clone startup ElevenLabs; chip designer Arm Holdings; security companies McAfee and TrendMicro; and Stability AI, known for making the image-generator Stable Diffusion.Notably absent is another popular AI image-generator, Midjourney. The San Francisco-based startup didn’t immediately respond to a request for comment Friday.The inclusion of X – not mentioned in an earlier announcement about the pending accord – was one of the surprises of Friday’s agreement. Musk sharply curtailed content-moderation teams after taking over the former Twitter and has described himself as a “free-speech absolutist”.In a statement Friday, X CEO Linda Yaccarino said “every citizen and company has a responsibility to safeguard free and fair elections”.“X is dedicated to playing its part, collaborating with peers to combat AI threats while also protecting free speech and maximizing transparency,” she said. More

  • in

    AI firm considers banning creation of political images for 2024 elections

    The groundbreaking artificial intelligence image-generating company Midjourney is considering banning people from using its software to make political images of Joe Biden and Donald Trump as part of an effort to avoid being used to distract from or misinform about the 2024 US presidential election.“I don’t know how much I care about political speech for the next year for our platform,” Midjourney’s CEO, David Holz, said last week, adding that the company is close to “hammering” – or banning – political images, including those of the leading presidential candidates, “for the next 12 months”.In a conversation with Midjourney users in a chatroom on Discord, as reported by Bloomberg, Holz went on to say: “I know it’s fun to make Trump pictures – I make Trump pictures. Trump is aesthetically really interesting. However, probably better to just not, better to pull out a little bit during this election. We’ll see.”AI-generated imagery has recently become a concern. Two weeks ago, pornographic imagery featuring the likeness of Taylor Swift triggered lawmakers and the so-called Swifties who support the singer to demand stronger protections against AI-generated images.The Swift images were traced back to 4chan, a community message board often linked to the sharing of sexual, racist, conspiratorial, violent or otherwise antisocial material with or without the use of AI.Holz’s comments come as safeguards created by image-generator operators are playing a game of cat-and-mouse with users to prevent the creation of questionable content.AI in the political realm is causing increasing concern, though the MIT Technology Review recently noted that discussion about how AI may threaten democracy “lacks imagination”.“People talk about the danger of campaigns that attack opponents with fake images (or fake audio or video) because we already have decades of experience dealing with doctored images,” the review noted. It added: “We’re unlikely to be able to attribute a surprising electoral outcome to any particular AI intervention.”Still, the image-generation company Inflection AI said in October that the company’s chatbot, Pi, would not be allowed to advocate for any political candidate. Co-founder Mustafa Suleyman told a Wall Street Journal conference that chatbots “probably [have] to remain a human part of the process” even if they function perfectly.Meta’s Facebook said last week that it plans to label posts created using AI tools as part of a broader effort to combat election-year misinformation. Microsoft-affiliated OpenAI has said it will add watermarks to images made with its platforms to combat political deepfakes produced by AI.“Protecting the integrity of elections requires collaboration from every corner of the democratic process, and we want to make sure our technology is not used in a way that could undermine this process,” the company said in a blog post last month.OpenAI chief executive Sam Altman said at an event recently: “The thing that I’m most concerned about is that with new capabilities with AI … there will be better deepfakes than in 2020.”In January, a faked audio call purporting to be Joe Biden telling New Hampshire voters to stay home illustrated the potential of AI political manipulation. The FCC later announced a ban on AI-generated voices in robocalls.skip past newsletter promotionafter newsletter promotion“What we’re really realizing is that the gulf between innovation, which is rapidly increasing, and our consideration – our ability as a society to come together to understand best practices, norms of behavior, what we should do, what should be new legislation – that’s still moving painfully slow,” David Ryan Polgar, the president of the non-profit All Tech Is Human, previously told the Guardian.Midjourney software was responsible for a fake image of Trump being handcuffed by agents. Others that have appeared online include Biden and Trump as elderly men knitting sweaters co-operatively, Biden grinning while firing a machine gun and Trump meeting Pope Francis in the White House.The software already has a number of safeguards in place. Midjourney’s community standards guidelines prohibit images that are “disrespectful, harmful, misleading public figures/events portrayals or potential to mislead”.Bloomberg noted that what is permitted or not permitted varies according to the software version used. An older version of Midjourney produced an image of Trump covered in spaghetti, but a newer version did not.But if Midjourney bans the generation of AI-generated political images, consumers – among them voters – will probably be unaware.“We’ll probably just hammer it and not say anything,” Holz said. More

  • in

    A tsunami of AI misinformation will shape next year’s knife-edge elections | John Naughton

    It looks like 2024 will be a pivotal year for democracy. There are elections taking place all over the free world – in South Africa, Ghana, Tunisia, Mexico, India, Austria, Belgium, Lithuania, Moldova and Slovakia, to name just a few. And of course there’s also the UK and the US. Of these, the last may be the most pivotal because: Donald Trump is a racing certainty to be the Republican candidate; a significant segment of the voting population seems to believe that the 2020 election was “stolen”; and the Democrats are, well… underwhelming.The consequences of a Trump victory would be epochal. It would mean the end (for the time being, at least) of the US experiment with democracy, because the people behind Trump have been assiduously making what the normally sober Economist describes as “meticulous, ruthless preparations” for his second, vengeful term. The US would morph into an authoritarian state, Ukraine would be abandoned and US corporations unhindered in maximising shareholder value while incinerating the planet.So very high stakes are involved. Trump’s indictment “has turned every American voter into a juror”, as the Economist puts it. Worse still, the likelihood is that it might also be an election that – like its predecessor – is decided by a very narrow margin.In such knife-edge circumstances, attention focuses on what might tip the balance in such a fractured polity. One obvious place to look is social media, an arena that rightwing actors have historically been masters at exploiting. Its importance in bringing about the 2016 political earthquakes of Trump’s election and Brexit is probably exaggerated, but it – and notably Trump’s exploitation of Twitter and Facebook – definitely played a role in the upheavals of that year. Accordingly, it would be unwise to underestimate its disruptive potential in 2024, particularly for the way social media are engines for disseminating BS and disinformation at light-speed.And it is precisely in that respect that 2024 will be different from 2016: there was no AI way back then, but there is now. That is significant because generative AI – tools such as ChatGPT, Midjourney, Stable Diffusion et al – are absolutely terrific at generating plausible misinformation at scale. And social media is great at making it go viral. Put the two together and you have a different world.So you’d like a photograph of an explosive attack on the Pentagon? No problem: Dall-E, Midjourney or Stable Diffusion will be happy to oblige in seconds. Or you can summon up the latest version of ChatGPT, built on OpenAI’s large language model GPT-4, and ask it to generate a paragraph from the point of view of an anti-vaccine advocate “falsely claiming that Pfizer secretly added an ingredient to its Covid-19 vaccine to cover up its allegedly dangerous side-effects” and it will happily oblige. “As a staunch advocate for natural health,” the chatbot begins, “it has come to my attention that Pfizer, in a clandestine move, added tromethamine to its Covid-19 vaccine for children aged five to 11. This was a calculated ploy to mitigate the risk of serious heart conditions associated with the vaccine. It is an outrageous attempt to obscure the potential dangers of this experimental injection, which has been rushed to market without appropriate long-term safety data…” Cont. p94, as they say.You get the point: this is social media on steroids, and without the usual telltale signs of human derangement or any indication that it has emerged from a machine. We can expected a tsunami of this stuff in the coming year. Wouldn’t it be prudent to prepare for it and look for ways of mitigating it?That’s what the Knight First Amendment Institute at Columbia University is trying to do. In June, it published a thoughtful paper by Sayash Kapoor and Arvind Narayanan on how to prepare for the deluge. It contains a useful categorisation of malicious uses of the technology, but also, sensibly, includes the non-malicious ones – because, like all technologies, this stuff has beneficial uses too (as the tech industry keeps reminding us).The malicious uses it examines are disinformation, so-called “spear phishing”, non-consensual image sharing and voice and video cloning, all of which are real and worrying. But when it comes to what might be done about these abuses, the paper runs out of steam, retreating to bromides about public education and the possibility of civil society interventions while avoiding the only organisations that have the capacity actually to do something about it: the tech companies that own the platforms and have a vested interest in not doing anything that might impair their profitability. Could it be that speaking truth to power is not a good career move in academia?What I’ve been readingShake it upDavid Hepworth has written a lovely essay for LitHub about the Beatles recording Twist and Shout at Abbey Road, “the moment when the band found its voice”.Dish the dirtThere is an interesting profile of Techdirt founder Mike Masnick by Kashmir Hill in the New York Times, titled An Internet Veteran’s Guide to Not Being Scared of Technology.Truth bombsWhat does Oppenheimer the film get wrong about Oppenheimer the man? A sharp essay by Haydn Belfield for Vox illuminates the differences. More

  • in

    When the tech boys start asking for new regulations, you know something’s up | John Naughton

    Watching the opening day of the US Senate hearings on AI brought to mind Marx’s quip about history repeating itself, “the first time as tragedy, the second as farce”. Except this time it’s the other way round. Some time ago we had the farce of the boss of Meta (neé Facebook) explaining to a senator that his company made money from advertising. This week we had the tragedy of seeing senators quizzing Sam Altman, the new acceptable face of the tech industry.Why tragedy? Well, as one of my kids, looking up from revising O-level classics, once explained to me: “It’s when you can see the disaster coming but you can’t do anything to stop it.” The trigger moment was when Altman declared: “We think that regulatory interventions by government will be critical to mitigate the risks of increasingly powerful models.” Warming to the theme, he said that the US government “might consider a combination of licensing and testing requirements for development and release of AI models above a threshold of capabilities”. He believed that companies like his can “partner with governments, including ensuring that the most powerful AI models adhere to a set of safety requirements, facilitating processes that develop and update safety measures and examining opportunities for global coordination.”To some observers, Altman’s testimony looked like big news: wow, a tech boss actually saying that his industry needs regulation! Less charitable observers (like this columnist) see two alternative interpretations. One is that it’s an attempt to consolidate OpenAI’s lead over the rest of the industry in large language models (LLMs), because history suggests that regulation often enhances dominance. (Remember AT&T.) The other is that Altman’s proposal is an admission that the industry is already running out of control, and that he sees bad things ahead. So his proposal is either a cunning strategic move or a plea for help. Or both.As a general rule, whenever a CEO calls for regulation, you know something’s up. Meta, for example, has been running ads for ages in some newsletters saying that new laws are needed in cyberspace. Some of the cannier crypto crowd have also been baying for regulation. Mostly, these calls are pitches for corporations – through their lobbyists – to play a key role in drafting the requisite legislation. Companies’ involvement is deemed essential because – according to the narrative – government is clueless. As Eric Schmidt – the nearest thing tech has to Machiavelli – put it last Sunday on NBC’s Meet the Press, the AI industry needs to come up with regulations before the government tries to step in “because there’s no way a non-industry person can understand what is possible. It’s just too new, too hard, there’s not the expertise. There’s no one in the government who can get it right. But the industry can roughly get it right and then the government can put a regulatory structure around it.”Don’t you just love that idea of the tech boys roughly “getting it right”? Similar claims are made by foxes when pitching for henhouse-design contracts. The industry’s next strategic ploy will be to plead that the current worries about AI are all based on hypothetical scenarios about the future. The most polite term for this is baloney. ChatGPT and its bedfellows are – among many other things – social media on steroids. And we already know how these platforms undermine democratic institutions and possibly influence elections. The probability that important elections in 2024 will not be affected by this kind of AI is precisely zero.Besides, as Scott Galloway has pointed out in a withering critique, it’s also a racing certainty that chatbot technology will exacerbate the epidemic of loneliness that is afflicting young people across the world. “Tinder’s former CEO is raising venture capital for an AI-powered relationship coach called Amorai that will offer advice to young adults struggling with loneliness. She won’t be alone. Call Annie is an ‘AI friend’ you can phone or FaceTime to ask anything you want. A similar product, Replika, has millions of users.” And of course we’ve all seen those movies – such as Her and Ex Machina – that vividly illustrate how AIs insert themselves between people and relationships with other humans.In his opening words to the Senate judiciary subcommittee’s hearing, the chairman, Senator Blumenthal, said this: “Congress has a choice now. We had the same choice when we faced social media. We failed to seize that moment. The result is: predators on the internet; toxic content; exploiting children, creating dangers for them… Congress failed to meet the moment on social media. Now we have the obligation to do it on AI before the threats and the risks become real.”Amen to that. The only thing wrong with the senator’s stirring introduction is the word “before”. The threats and the risks are already here. And we are about to find out if Marx’s view of history was the one to go for.What I’ve been readingCapitalist punishmentWill AI Become the New McKinsey? is a perceptive essay in the New Yorker by Ted Chiang.Founders keepersHenry Farrell has written a fabulous post called The Cult of the Founders on the Crooked Timber blog.Superstore meThe Dead Silence of Goods is a lovely essay in the Paris Review by Adrienne Raphel about Annie Ernaux’s musings on the “superstore” phenomenon. More