Ofcom called out Twitch, which is owned by Amazon. (Gabby Jones/Bloomberg News)
The UK’s communications regulator, Ofcom, reported findings from its first year overseeing harmful content on platforms like TikTok, Snap, Twitch and OnlyFans, and the debut report may hold some lessons for U.S. lawmakers.
Ofcom, which has authority over TV, telecom, and radio, began regulating some online video platforms in October 2021, under a set of expanded rules on harmful content introduced as a precursor to the controversial Online Safety Bill, a sweeping proposal that sought to police “legal, but harmful” content online.
Before her resignation, UK Prime Minister Liz Truss made limited remarks about the bill, voicing concerns about child safety online and protecting free speech.
The rules only applied to video content, focusing on protecting minors from harmful content and protecting all users from hate speech and incitements to violence, as well as content around terrorism, racism and child sexual abuse material, which can be a criminal offense in the UK.
Platforms self-assess whether the rules for video-sharing platforms apply to them. Instagram, Facebook and Twitter are not in the report.
In its report, Ofcom called out the streaming video service Twitch, which is owned by Amazon, for lacking safeguards to prevent minors from watching content that creators have marked as mature, even when the user declares themselves under 18. (Amazon founder Jeff Bezos owns The Washington Post.)
In a statement, Twitch spokesperson Samantha Faught said the company has developed safety tools and policies unique to live video streaming. “We hold all of our users accountable to a robust set of global Community Guidelines which — among others — prevent nudity, sexually suggestive content, or extreme violence and are enforced through proactive technologies and industry-leading operations for 24/7/365 review of user reports,” Fought wrote. “We are always working to enhance protections for our community, and have significant work in progress that will specifically seek to further protect users aged 13-17.”
The report also identified issues with Snapchat owner Snap, noting that it was “not clear” whether Snap had a formalized structure to govern safety and that Snap’s decision-making structure to assess safety risks was also “not clear.”
Ofcom’s findings on internal company structure and decision-making were based on information provided by the platform. According to Snap’s self-assessment, only two features were within the scope of the video-sharing rules: Spotlight, which shows user-generated content that goes through human moderation before exceeding 25 views, and Discover, which features curated content from verified media outlets and approved content creators.
In a statement, Snap’s director of platform policy, Nona Farahnik, described an established process for assessing online risks that seems to differ from Ofcom’s findings. “At Snap, we do have a formal Safety Governance Team,” Farahnik wrote. “This team is in constant communication to implement our platform safety governance framework [and] use our long-held safety and privacy by design processes to vet new features in development.”
Policing social media companies like they are broadcasters is an astute and effective approach that reflects the way these platforms operate, said Imran Ahmed, founder and CEO of the Center for Countering Digital Hate, a nonprofit advocacy group that focuses on hate speech and disinformation online. Video is hard to search through, he said, and impossible for groups like his to monitor without government oversight.
But while the report identified some key harms, its piecemeal findings underscored the lack of transparency when it comes to the way these platforms operate.
“What underpins effective regulation is transparency. The UK has kind of skipped a step,” he said.
One way to achieve that level of transparency might be to enforce UK standards not just on whether a platform is abiding by just the letter of the law, but whether it’s abiding by the spirit of the law, Ahmed said. He pointed to the trove of documents released by Facebook whistleblower Frances Haugen that showed the company knew the staggeringly low percentage of hate speech removed in India, the company’s largest market, despite touting rosier numbers in its annual reports.
“What you still don’t have is platforms forming meaningful partnerships with regulators to help collectively minimize harms,” Ahmed said. Nonetheless, he noted, that two areas of bipartisan consensus in the House are around protecting children online and algorithmic transparency. “When [companies] get a complaint, what rule they assess against, what conclusions they come to, and in whose interest,” he explained.
At the very least, the report shows the virtue of action, Ahmed argued. “Effective regulation is an iterative process, you have to be willing to try and to learn and to enter a dialogue,” he said.
Director of the Tech Transparency Project, a nonprofit tech watchdog that receives funding from the George Soros-founded Open Society Foundations, Omidyar Network and others, Katie Paul said in a statement that she hoped US regulators would follow Ofcom’s example of focusing on Big Tech’s harms to children. “As a society, we are running an unprecedented, mass experiment on our children’s developing brains with almost no research or guidance from the government.”
Director of the digital privacy group Fight for the Future, Evan Greer sees it differently. “There’s no global consensus on what content counts as ‘inappropriate’ for young people. Platforms end up making judgment calls, and we’ve seen in the past that they often deem LGBTQ+ related content, or content about consent, sexual health, or other important information, as ‘mature,’ often making it inaccessible to young people who need it most,” Greer said.
“Once you get the government involved in mandating that platforms take certain steps around content moderation, there’s even more room for abuse,” Greer added. “If, for example, we had a similar law in the US, it’s not hard to imagine how an attorney general like [Texas’s] Ken Paxton would define ‘content that’s harmful to kids.’”
Our top tabs
US government asked Mexico not to buy Chinese technology for border
The letter from US Ambassador to Mexico Ken Salazar warned that bilateral cooperation between Mexico and the United States could be hampered by use of the equipment.
US Ambassador to Mexico Ken Salazar’s letter to Mexican Foreign Minister Marcelo Ebrard urged Mexican authorities not to buy equipment from Beijing-based Nuctech, which US officials worried would give China access to data about goods entering the United States, Kevin Sieff and Nick Miroff report. In the previously unreported May 2 letter, Salazar wrote that US-Mexico bilateral cooperation “could be put at risk by the use of unreliable equipment.”
The letter was one of the millions of documents that hackers leaked after targeting Mexico’s secretariat of defense. Internal documents show that Mexico already began purchasing Nuctech scanners even before May 2.
Civil society organization Mexicans Against Corruption shared the documents with The Post. The Post independently verified the documents.
The Department of Homeland Security said in 2020 that “Nuctech very likely has a close and enduring relationship with the Chinese Government to advance Nuctech’s business interests and develop screening and detection systems on behalf of the Chinese Government.” It also wrote that the company’s equipment likely has “deficiencies in detection capabilities, which may create opportunities for exploitation by the Chinese Government.”
Nuctech, which didn’t respond to requests for comment, said in an undated statement on its website that “it is not state controlled” and its customers are the “sole owners of all data generated by Nuctech’s systems.” The firm is “100 percent committed to the safety and security of our customers and their data and any suggestions to the contrary is categorically false and designed to stifle emerging market competition,” it said.
Indian news outlet retracts stories on Meta
The Wire’s retractions come around two weeks after the Indian news outlet published an explosive story alleging that Facebook parent Meta gave an official from India’s ruling party the ability to censor posts. Pressure had been growing on the outlet after experts pointed to a growing list of discrepancies in videos and emails that the Wire presented as proof of its allegations, my colleagues reported last week.
The Wire said that its ongoing investigation of the stories “does not as yet allow us to take a conclusive view about the authenticity and bona fides of the sources with whom a member of our reporting team says he has been in touch over an extended period of time.” But in light of the discrepancies, it said, “The Wire believes it is appropriate to retract the stories.”
RNC sues Google over spam filtering
The RNC isn’t in Google’s pilot program, which lets candidates and campaigns apply to bypass spam filters. (David Gray/Bloomberg News)
The Republican National Committee said in its lawsuit that Google is “unlawfully discriminating” against it by putting its emails in spam folders, Axios’s Sara Fischer and Ashley Gold report. The RNC isn’t enrolled in Google’s pilot program, which lets some political candidates and campaigns bypass its spam filters, Axios reported.
The RNC has also “made few efforts to alter the core practices that might result in their emails being labeled as spam,” the Verge’s Makena Kelly wrote this month.
Democrats initially hammered Google for launching the program, but the Democratic National Committee has joined, Cristiano Lima reported last week. The DNC still objects to Google’s launching of the program but won’t use it to send Gmail users scams and fraud, DNC spokesperson Daniel Wessel told The Technology 202 at the time.
Google has denied the RNC’s allegations. “As we have repeatedly said, we simply don’t filter emails based on political affiliation,” Google spokesperson José Castañeda told Axios. “Gmail’s spam filters reflect users’ actions. We provide training and guidelines to campaigns, we recently launched an FEC-approved pilot for political senders, and we continue to work to maximize email deliverability while minimizing unwanted spam.” https://www.washingtonpost.com
