Content Policy

X moves to comply with Brazil’s court amid misinformation crackdown

After months of defiance, Elon Musk’s social media platform, X, has told Brazil’s Supreme Court that it has complied with orders to curb the spread of misinformation. The direction shift comes as Musk seeks to lift a ban on the platform following a prolonged battle with the Brazilian judiciary over what he called ‘censorship.’ The court had suspended access to X in late August, leaving Brazilians needing the platform in one of its largest global markets.

The dispute revolves around Justice Alexandre de Moraes, who has been leading a crackdown on what he perceives as threats to democracy and the political use of disinformation. After Musk closed X’s office in Brazil, the judge banned the platform. Moraes also froze accounts of Musk’s satellite company, Starlink, prompting Musk to label him a ‘dictator.’

X, which boasts 21.5 million users in Brazil, attempted to circumvent the ban by using third-party cloud services, allowing temporary access. However, the effort was short-lived, especially after Moraes threatened heavy fines. In a more conciliatory move, X recently appointed a local legal representative in Brazil, signalling a shift in its approach.

In documents submitted to the court, X confirmed that it had blocked nine accounts linked to a hate speech and misinformation investigation. The action reflects a change in Musk’s strategy, as sources close to him suggest he now plans to comply with local laws while continuing to defend free speech through legal avenues.

Why does it matter?

Musk’s battle with the Brazilian judiciary mirrors similar tensions in countries like Australia and the UK, where governments are working to curb online misinformation. Despite his stance as a free speech advocate, Musk’s recent actions in Brazil indicate a more pragmatic approach to navigating regulatory challenges in key markets.

As X awaits the court’s decision, which could restore access within days, analysts believe that Musk’s surprising backtracking demonstrates a recognition that fighting the law in Brazil could have further damaged his standing in the country. Now, the company appears ready to respect legal boundaries, even if it means fighting battles in the courtroom rather than on the platform.

FTC fines companies for misusing AI in e-commerce schemes

The US Federal Trade Commission (FTC) has cracked down on five companies for deceptive use of AI. Three cases involved businesses falsely claiming to help consumers generate passive income through e-commerce. The FTC also reached settlements with DoNotPay and Rytr, two companies accused of misleading consumers with their AI tools. DoNotPay, which marketed automated legal services, agreed to a $193,000 settlement and will notify customers of the tool’s limitations, while Rytr faced criticism for allowing users to create fake product reviews through its AI writing feature.

FTC Chair Lina M. Khan stressed that AI tools must comply with existing laws, making it clear that deceiving or misleading consumers with AI is illegal. Despite not admitting wrongdoing, both Rytr and DoNotPay settled with the FTC. Rytr agreed to discontinue its review-generating feature, used to create fake product reviews, while DoNotPay accepted a settlement without admitting fault.

The FTC’s actions have sparked internal debate on how to regulate AI. While all five commissioners supported cracking down on false AI claims, the two Republican commissioners raised concerns about the agency’s authority in the Rytr case. This division highlights differing views within the FTC on the scope of its regulatory powers when addressing AI-related issues.

Judge rules NYC food delivery data law unconstitutional

A federal judge has ruled that New York City’s law requiring food delivery companies to share customer data with restaurants is unconstitutional. The decision, handed down by US District Judge Analisa Torres, found the law violated the First Amendment by regulating commercial speech inappropriately.

The law, introduced in 2021 to support local restaurants recovering from the COVID-19 pandemic, required delivery platforms like DoorDash and UberEats to share customer details. Delivery companies in US argued that the law threatened both customer privacy and their business by allowing restaurants to use the data for their own marketing purposes.

Judge Torres stated that New York City failed to prove the law was necessary and suggested alternative methods to support restaurants, such as letting customers opt-in to share their data or providing financial incentives. City officials are reviewing the ruling, while delivery companies hailed it as a victory for data protection.

The New York City Hospitality Alliance expressed disappointment, claiming the ruling hurts small businesses and calling for the city to appeal the decision.

AI-powered translations now available on Reddit globally

Reddit is expanding its AI-powered translation feature to over 35 countries in Europe, Asia, and Latin America. This marks a significant shift in the platform’s approach to increasing accessibility for non-English-speaking users. Reddit’s latest move follows its earlier translation service for French, broadening its language offerings to accommodate its growing user base.

The new feature enables automatic translation for both posts and comments, allowing conversations across multiple languages without needing manual input. Users can choose their preferred language settings, facilitating smoother interactions in diverse communities. Development like this one allows multilingual conversations to take place seamlessly within a single thread.

Reddit’s AI-powered translations extend beyond user interactions. Search engines will also index translated content, potentially drawing in more users who are searching for content in their local language. Reddit has already rolled out support in Brazil and Spain, with plans to expand to Germany, Italy, the Philippines, and other regions in Latin America soon.

The company continues to see growth in both its user base and ad revenue. Enhancing language support with AI-powered tools is a strategic move to appeal to global audiences, positioning Reddit for further international expansion.

Microsoft unveils tool to fix AI hallucinations amid expert concerns

Microsoft has introduced a new service, called Correction, aimed at addressing a significant flaw in AI models, hallucinations, or factually incorrect responses. The tool identifies and revises erroneous AI-generated content by cross-referencing with accurate data sources, such as transcripts. Correction, available through Microsoft’s Azure AI Content Safety API, works with various models, including OpenAI’s GPT-4.

While Microsoft promotes Correction as a way to boost AI reliability, experts remain skeptical. Researchers warn that hallucinations are deeply ingrained in how AI models operate. Since these systems rely on statistical patterns rather than actual knowledge, completely eliminating false outputs might be impossible. They also caution that this solution could create new issues, like giving users a false sense of trust in AI outputs.

Despite these concerns, Microsoft is pushing to demonstrate the value of its AI tools, having invested billions in the technology. However, concerns about performance and cost are mounting, with some clients already pausing AI deployments due to inaccuracies and high expenses. Experts argue that AI, still in its developmental stages, is being rushed into industries without fully addressing its flaws.

UN adopts ‘Pact for the Future’

On 22 September 2024, world leaders convened in New York to adopt the ‘Pact for the Future’ – a comprehensive agreement designed to reimagine global governance in response to contemporary and future challenges.

The ground-breaking Pact includes a Global Digital Compact and a Declaration on Future Generations, aiming to update the international system established by previous generations. The Secretary-General stressed the importance of aligning global governance structures with the realities of today’s world, fostering a more inclusive and representative international system.

The Pact covers many critical areas, including peace and security, sustainable development, climate change, digital cooperation, human rights, and gender equality. It marks a renewed multilateral commitment to nuclear disarmament and advocates for strengthened international frameworks to govern outer space and prevent the misuse of new technologies. To bolster sustainable development, the Pact aims to accelerate the Sustainable Development Goals (SDGs), reform international financial architecture, and enhance measures to tackle climate change by committing to net-zero emissions by 2050.

Digital cooperation is notably addressed through the Global Digital Compact, which outlines commitments to connect all people to the internet, safeguard online spaces, and govern AI. The Compact promotes open-source data and sets the stage for global data governance. It also ensures increased investment in digital public goods and infrastructure, especially in developing countries.

Why does it matter?

The ‘Pact for the Future’ encapsulates a detailed, optimistic vision geared toward creating a sustainable, just, and peaceful global order. The Summit of the Future, which facilitated the adoption of this Pact as an extensively inclusive process, involves millions of voices and contributions from diverse stakeholders. The event was attended by over 4,000 participants, including global leaders and representatives from various sectors, and was preceded by Action Days, which drew more than 7,000 attendees. Such a forum shows firm global commitments to action, including pledges amounting to USD 1.05 billion to advance digital inclusion.

Snapchat’s balance between user safety and growth remains a challenge

Snapchat is positioning itself as a healthier social media alternative for teens, with CEO Evan Spiegel emphasising the platform’s different approach at the company’s annual conference. Recent research from the University of Amsterdam supports this view, showing that while platforms like TikTok and Instagram negatively affect youth mental health, Snapchat use appears to have positive effects on friendships and well-being.

However, critics argue that Snapchat’s disappearing messages feature can facilitate illegal activities. Matthew Bergman, an advocate for social media victims, claimed the platform has been used by drug dealers, citing instances of children dying from fentanyl poisoning after buying drugs via the app. Despite these concerns, Snapchat remains popular, particularly with younger users.

Industry analysts recognise the platform’s efforts but highlight its ongoing challenges. As Snapchat continues to grow its user base, balancing privacy and safety with revenue generation remains a key issue, especially as it struggles to compete with bigger players like TikTok, Meta, and Google for advertising.

Snapchat’s appeal lies in its low-pressure environment, with features like disappearing stories and augmented reality filters. Young users, like 14-year-old Lily, appreciate the casual nature of communication on the platform, while content creators praise its ability to offer more freedom and reduce social pressure compared to other social media platforms.

Brazil’s betting boom raises economic concerns

Brazil is witnessing a rapid rise in online sports betting, attracting foreign companies but raising alarms over its economic impact. While the government anticipates increased tax revenue, experts are concerned that gambling diverts money from consumer spending. Gabriel Galipolo, the incoming central bank governor, noted that this surge may hinder the benefits of rising incomes, as savings and consumption growth appear to be stalling due to the increased focus on betting.

Brazilians spent more than $12 billion on foreign betting platforms in the past year, placing the country among the world’s largest sports betting markets. Despite laws prohibiting the use of credit cards for betting, critics warn that gambling is taking a toll on household budgets, particularly among lower-income families. An increasing share of family income is being funnelled into betting, reducing spending on essentials like food, clothing, and healthcare.

The rapid growth of the gambling industry in Brazil has attracted major players like Betfair, Betsson, and Caesars Sportsbook, all keen to tap into the country’s 200 million sports enthusiasts. However, research from the U.S. indicates that legalised betting can lead to serious long-term consequences, such as rising credit card debt, bankruptcies, and economic instability for struggling families. Brazil now faces the challenge of balancing the potential economic benefits of the betting sector with the risks of increasing financial pressure on its citizens.

Social media owners, politicians, and governments top threats to online news trust, IPIE report shows

A recent report from the International Panel on the Information Environment (IPIE) highlights social media owners, politicians, and governments as the primary threats to a trustworthy online news landscape. The report surveyed 412 experts across various academic fields and warned of the unchecked power social media platforms wield over content distribution and moderation. According to Philip Howard, a panel co-founder, such results pose a critical threat to the global flow of reliable information.

The report also raised concerns about major platforms like X (formerly Twitter), Facebook, Instagram, and TikTok. Allegations surfaced regarding biassed moderation, with Elon Musk’s X reportedly prioritising the owner’s posts and Meta being accused of neglecting non-English content. TikTok, under scrutiny for potential ties to the Chinese government, has consistently denied being an agent of any country. The panel emphasised that these platforms’ control over information significantly impacts public trust.

The survey revealed that around two-thirds of respondents anticipate the information environment will deteriorate, marking a noticeable increase in concern compared to previous years. Experts cited AI tools as a growing threat, with generative AI exacerbating the spread of misinformation. AI-generated videos and voice manipulation ranked as the top concerns, especially in developing countries with more acute impact.

However, not all views on AI are negative. Most respondents also saw its potential to combat misinformation by helping journalists sift through large datasets and detect false information. The report concluded by suggesting key solutions: promoting independent media, launching digital literacy initiatives, and enhancing fact-checking efforts to mitigate the negative trends in the digital information landscape.

Telegram to share user data with authorities

Telegram apparently decided to alleviate its policy restrictions and to provide users’ IP addresses and phone numbers to authorities in response to valid legal requests. The shift in policy, announced by CEO Pavel Durov, marks a significant change for the platform, which has long been known for its resistance to government data demands. The update comes in the wake of Durov’s recent legal troubles in France, where he is facing charges related to the spread of child abuse materials on the platform.

Durov, under investigation since his arrest in France last month, says the new measures are part of broader efforts to deter criminal activity on Telegram. Historically, Telegram has been criticised for its lax approach to moderation, often ignoring government requests to remove illegal content or share information on suspected criminals. Now, with AI and human moderators, the app conceals problematic content from search results.

The case against Durov has intensified scrutiny of Telegram’s role in facilitating illegal activities. French authorities have accused Durov of refusing to cooperate with law enforcement by not providing data for wiretaps related to criminal investigations. Durov denies the charges despite these accusations and has remained in France in the inquiry.

Why does this matter?

Telegram has long been a tool for activists and dissidents, especially in countries like Russia and Iran, where it has been used to challenge authoritarian regimes. However, the platform has also attracted extremists, conspiracy theorists, and white supremacists. In some cases, Telegram has been used to coordinate real-world attacks, leading to mounting pressure on the company to take greater responsibility.

In response to these challenges, Telegram has introduced several policy changes. Earlier this month, the platform disabled new media uploads to combat bots and scammers. These moves signal a new chapter for Telegram as it navigates the delicate balance between privacy, free speech, and public safety.