How can people fight back against AI deepfakes? More AI, experts say
The World Intellectual Property Organisation (WIPO) defines a deepfake as an AI technique that synthesises media by either superimposing human features on another body or manipulating sounds to generate a realistic video.
This year, high-profile deepfake scams have targeted US Secretary of State Marco Rubio, Italian defense minister Guido Crosetto, and several celebrities, including Taylor Swift and Joe Rogan, whose voices were used to promote a scam that promised people government funds.
Deepfakes were created every five minutes in 2024, according to a recent report from think tank Entrust Cybersecurity Institute.
What impacts do deepfakes have?
Deepfakes can have serious consequences, like the disclosure of sensitive information with government officials who sound like Rubio or Crosetto.
'You're either trying to extract sensitive secrets or competitive information or you're going after access, to an email server or other sensitive network,' Kinny Chan, CEO of the cybersecurity firm QiD, said of the possible motivations.
Synthetic media can also aim to alter behaviour, like a scam that used the voice of then-US President Joe Biden to convince voters not to participate in their state's elections last year.
"While deepfakes have applications in entertainment and creativity, their potential for spreading fake news, creating non-consensual content and undermining trust in digital media is problematic," the European Parliament wrote in a research briefing.
The European Parliament predicted that 8 million deepfakes will be shared throughout the European Union this year, up from 500,000 in 2023.
What are some ways AI is fighting back?
AI tools can be trained through binary classification so they can classify data being fed into them as being real or fake.
For example, researchers at the University of Luxembourg said they presented AI with a series of images with either a real or a fake tag on them so that the model gradually learned to recognise patterns in fake images.
'Our research found that ... we could focus on teaching them to look for real data only,' researcher Enjie Ghorbel said. 'If the data examined doesn't align with the patterns of real data, it means that it's fake".
Another solution proposed by Vijay Balasubramaniyan, CEO and founder of the tech firm Pindrop Security, is a system that analyses millions of data points in any person's speech to quickly identify irregularities.
The system can be used during job interviews or other video conferences to detect if the person is using voice cloning software, for instance.
Someday, deepfakes may go the way of email spam, a technological challenge that once threatened to upend the usefulness of email, said Balasubramaniyan, Pindrop's CEO.
'You can take the defeatist view and say we're going to be subservient to disinformation,' he said. 'But that's not going to happen".
The EU AI Act, which comes into force on August 1, requires that all AI-generated content, including deepfakes, are labelled so that users know when they come across fake content online.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Euronews
a day ago
- Euronews
Why women politicians face more online hate
Giulia Fossati entered Italian politics around 2021, turning regularly to social media to share her views on topics like migration, racism, and feminism. But her online presence came at a cost. 'There is great violence on social media,' said Fossati, a member of the centre-left Partito Democratico who represents women who are registered party members in Pavia, near Milan. 'I get many comments, especially when I talk about feminist topics,' she told Euronews Next, citing examples like 'go to the kitchen,' or 'idiot shut up'. Fossati is not yet a household name in Italian politics, yet she is already facing online harassment, with insults often combining digs at her gender and age. 'They call me a 'young woman' in a way that makes me sound less credible, less defensible than an adult,' she said. Fossati's experience is not an exception. Women politicians are more likely than their male counterparts to face identity-based attacks on social media, according to a new peer-reviewed study published in the journal Politics and Gender. Researchers analysed more than 23 million posts on the social media platform X, formerly known as Twitter, that were addressed to politicians in Germany, Spain, the United Kingdom, and the United States. At the time, the platform still had active content moderation. While men and women face a similar number of online attacks, male politicians tend to be targeted with general insults and female politicians are more often attacked for their appearance, gender, ethnicity, or personal morality, the study found. In Europe, fame has little to do with the attacks. Female politicians face uncivil tweets regardless of how well-known they are, and they are more exposed to such attacks than their male counterparts, the researchers found. The study defined 'uncivil' tweets as those containing hate speech, gender stereotypes, exclusionary language (like 'women should stay at home rather than do politics'), threats to individual rights, name-calling, character attacks ('liar,' 'traitor'), vulgarity, sarcasm, all-caps shouting, or content that is incendiary or humiliating. These online attacks, researchers warned, can lead women to reduce their presence online and deter them from running for political office. The deep-rooted causes of online hate towards women The study has some limitations. Andrea Pető, a professor in the department of gender studies at the Central European University in Vienna, criticised the study's use of AI, saying that while these models can flag explicit threats, they struggle to detect more subtle forms of verbal aggression. 'Artificial intelligence cannot catch the nuances,' Pető told Euronews Next. Likewise, by flagging comments as 'uncivil,' some context may be missed, for example, the fact that many 'democratic voters hold these supposedly 'uncivil' viewpoints,' she said. Even so, the study's overall conclusions came as no surprise to gender and politics experts. Online harassment of women has long been under scrutiny, prompting research, debate, and legal reform. Power, politics, and public debate have not historically been associated with female roles or traditions. Consider women's right to vote. In some European countries, such as Greece, universal suffrage for national elections was established only in 1952. The legacy of this gender inequality is evident even today. When women enter political spaces, including on social media, they may encounter hostility and attacks because they are women, Pető said. 'Women are expected to be in the private sphere and those questioning this divide, let they be witches, or Marie Curie or local politicians, or an MP, face a certain kind of disciplining action from the public sphere, run by men,' she said. But is this online hostility driven by deep-rooted societal attitudes, by the technological systems that amplify them, or both? Technological and economic issues 'Technology often works as a mirror,' said Sandra Wachter, professor of technology and regulation at the University of Oxford and at the Hasso Plattner Institute in Potsdam, Germany. 'Those who already experience oppression and discrimination in society face it on a larger scale if we implement technology in a completely unfettered way. And this is why law is important,' she told Euronews Next. Wachter noted that, beyond social and historical causes, online attacks against women are also driven by the economic interests of major tech companies. She said their business models are designed to keep users online as long as possible to sell advertisements. 'What people want to see and what keeps them engaged is something that is raging, outrageous,' Wachter said. That's a key reason why fake news, often characterised with a sensational tone, tends to diffuse farther and faster than legitimate information. Even so, many people are unaware of the problem, Wachter said. Victims of online attacks are often blamed, while perpetrators – and even law enforcement – frequently fail to grasp how serious the consequences can be, partly due to the digital setting, she said. How to fix the problem Some platforms, such as TikTok, use AI-driven content moderation, while others, such as Meta's Facebook and Instagram, have scaled back content reviews. But AI cannot catch everything, warned Sara de Vuyst, a professor in contemporary visual culture at the University of Maastricht in the Netherlands. 'This [the use of AI] has some issues; they're missing things when the comments are formulated in a more sarcastic way, nuances get lost,' de Vuyst told Euronews Next. Both de Vuyst and Wachter agree that regulations like the European Union's Digital Service Act (DSA) are a step in the right direction. Entered into force in February 2024, the DSA aims to protect consumers' rights online. It makes it easier for users to flag if an online post is problematic, and requires big social media companies to implement risk-reduction protocols. Yet both de Vuyst and Wachter argue that under the DSA, corporate accountability of these companies remains low. 'Those are all great, fantastic steps in the right direction,' Wachter said. 'But the thing that nobody has yet done is ask the question: 'what about the business model?'' Back in Italy, Fossati has taken matters into her own hands. At first, she spent time engaging in discussions with her online haters, trying to understand their perspectives. However, she adopted a different approach after realising that many commenters were not open to genuine debate. 'If someone insults me, my answer is always a very ironic one,' she said. If a comment is particularly offensive, she reminds the user that she could press charges, yet she has never done so because it would be an expensive and cumbersome process. Despite the attacks, Fossati keeps up her motivation and chooses to focus on the bright side. 'There are often negative comments because we don't write about how good people are,' she said. Online haters 'don't represent the whole reality'.
LeMonde
3 days ago
- LeMonde
AI Act: 38 global creators' organizations condemn 'betrayal' of Europe's stated goals
A bitter taste and deep disappointment are being felt among authors, performers, publishers, producers and other rights holders across Europe and around the world. In a joint statement signed on Thursday, July 31, 38 leading global rights holders' organizations expressed their "dissatisfaction" with the implementation of the European Union's regulation on artificial intelligence (AI). Their discontent applies to the "Code of Practice, the GPAI Guidelines, and the Template for disclosure of a sufficiently detailed summary of training data under Article 53 of the EU AI Act." This legislation regulating AI within the European Union partially takes effect on Saturday, August 2. Despite their extensive engagments made "throughout this process, the final outcomes fail to address the core concerns which our sectors [...] have consistently raised," the organizations state. The coalition went further: "The result is not a balanced compromise." They call it "a missed opportunity to provide meaningful protection of intellectual property rights in the context of GenAI [generative AI]." Worse still, the AI Act "does not deliver on the promise of the EU AI Act itself."
Euronews
3 days ago
- Euronews
Cruel summer turns up heat on Ursula von der Leyen's second mandate
It's fair to assume Ursula von der Leyen will be looking forward to her summer break. This July, typically a month of low-intensity in Brussels politics, has been nothing short of a whirlwind for the president of the European Commission, with consequential decisions and pivotal moments that could reshape the trajectory of her five-year mandate. Nobody expected her second term to be an easy ride, certainly not after the electoral victory of Donald Trump, a man whose beliefs are directly at odds with the bloc's defence of predictable rules, open markets and international cooperation. Still, the events of the last five weeks, a powerful blend of domestic bickering, global turmoil and personal scrutiny, crack the president's tightly controlled image and leave her vulnerable to a sort of stinging criticism she had previously avoided. Here's how von der Leyen's summer got crueller and bleaker. First, the motion Von der Leyen never enjoyed hugely harmonious relations with the European Parliament. MEPs have routinely complained about the president's well-known preference for engaging with member states, the real holders of political power, and her perceived tendency to treat the hemicycle as a second-rate legislator. Tensions and discontent had been simmering for months when a hard-right lawmaker, Romania's Gheorghe Piperea, drafted a motion of censure against the European Commission and managed to secure the necessary 72 signatures to put it to a vote. Piperea's motion, which combined the Pfizergate scandal with conspiracies about electoral interference, never had a realistic chance of succeeding. The far-fetched move was ultimately rejected with 360 votes against and 175 in favour. But the arithmetic was not the point. The motion put von der Leyen in a rare position of defiance. The Commission chief was forced to address, one by one, the accusations that Piperea had levelled against her, rejecting them all as "false claims" and "sinister plots". Socialists, liberals and greens, all of whom backed her re-election last year, seized the moment to air their pent-up frustration and run through a shopping list of recriminations, raising serious questions about the viability of the centrist coalition. "I will always be ready to debate any issue that this house wants, with facts and with arguments," she said, offering an olive branch for "unity". The saga polarised the Parliament and weakened von der Leyen. Crucially, it proved how relatively easy it is for MEPs to file a motion of censure at any point. Manon Aubry, the co-leader of The Left, has begun collecting signatures for a fresh attempt. Then, the budget Bruised from the motion of censure, von der Leyen shifted gears to focus on what was expected to be her biggest announcement of the year: the Commission's long-awaited proposal for the bloc's next seven-year budget (2028-2034). It was the perfect opportunity for von der Leyen to showcase her political gravitas, reframe the conversation and turn a page on the acrimonious vote. As it happened, the proposal was marred by internal fights over the total size of the budget, the restructuring of programmes and the financial allocation for each priority. Her novel idea to merge agricultural and cohesion funds into a single envelope leaked in advance and prompted immediate criticism from the powerful farming lobby. Her cabinet's penchant for secrecy left other Commissioners in a scramble to figure out how much money they would have in future for their portfolios. By the time von der Leyen unveiled the €2 trillion budget, the largest ever put forward, attention was split between her ground-breaking blueprint and the behind-the-scenes drama, which stretched through the night until the final meeting. During the press conference, the president was asked the awkward question on whether she had treated her 26 Commissioners with fairness and respect. "Not everyone was satisfied," she said, explaining the one-by-one consultations. "There's strong support. The collegial decision is taken. And now we have to fight to bring this budget further in the next two years." Later, the summit "Unsustainable." That is how Commission officials had described the state of EU-China relations in anticipation of a high-stakes bilateral summit in Beijing. China's generous use of state subsidies to boost domestic production despite lacking the internal demand to absorb it has provoked the fury of Brussels, which fears the intense race-to-the-bottom could decimate European industry. Beijing's decision to curb exports of critical raw materials, hinder market access for foreign firms and continue its "no-limits partnership" with Moscow added to the piled-up tensions. Despite the urgent need for tangible change, Ursula von der Leyen left the summit with little to show. There was a new commitment to address bottlenecks in the supply of rare earths and a joint statement on climate action. Beyond that, no progress was achieved, and the main points of friction were left conspicuously unaddressed. "We have reached a clear inflection point," von der Leyen told reporters. "As we said to the Chinese leadership, for trade to remain mutually beneficial, it must become more balanced. Europe welcomes competition. But it must be fair." The underwhelming summit suggests EU-China relations will remain confrontational for the foreseeable future, trapping von der Leyen between two perilous avenues: retaliate and risk facing Beijing's wrath or offer concessions that might not be reciprocated. "With its rare earth controls, China has given Europe a glimpse of the havoc it can wreak if the trade battle gets hot," Noah Barkin, a senior fellow at the German Marshall Fund, wrote in his latest newsletter. "But if Europe fails to push back forcefully, throwing all the defensive trade tools it has at China, the long-term damage to its industrial base is likely to be profound." And finally, the deal Ursula von der Leyen's admiration for the transatlantic alliance faced its most gruelling test on 2 April 2025, when Donald Trump unveiled his contentious "reciprocal" tariffs to single-handedly redesign the economic order built at the end of World War II. That fateful day triggered frantic negotiations to spare the export-oriented bloc from Trump's sweeping duties. His ultimatum to apply an across-the-board 30% rate, made in a letter addressed to von der Leyen, caused palpable panic across Brussels. With the deadline of 1 August looming ever closer, the Commission chief flew to Scotland and met Trump in a last-ditch attempt to seal a deal of sorts. What emerged from those talks was an agreement to apply a 15% tariff on the majority of EU products and a 0% tariff on the majority of US products. Additionally, the bloc made tentative pledges to spend an astonishing $750 billion on American energy and invest $600 billion in the American market by the end of Trump's mandate. The outcry was loud and fast: critics spoke of capitulation, humiliation and submission to decry the extremely lopsided nature of the deal, which codifies the highest tariffs that transatlantic commerce has seen in over 70 years. Von der Leyen, who had just stood firm against Beijing's demands, struggled to explain why she had offered such far-reaching concessions to satisfy Trump. "15% is not to be underestimated, but it is the best we could get," she said. The deal, factually disadvantageous for the bloc, takes the shine off von der Leyen's reputation as a reliable manager-in-chief and threatens to become a painful thorn in her second term, which is meant to prioritise competitiveness and growth. If anything, she might take comfort in the fact that none of the 27 EU leaders appear to have the stomach to tear the deal apart and start negotiations from scratch. "Europe does not yet see itself as a power," said French President Emmanuel Macron. "To be free, you must be feared. We were not feared enough."
