Google's autocomplete subtitles mislead public about conspiracy theorists
A study by researchers at Simon Fraser University found that Google's autocomplete subtitle algorithm assigned neutral or positive labels to 30 of 37 conspiracy theorists, including Alex Jones as 'American Radio Host' and Gavin McInnes as 'Canadian writer'. The researchers argue this can mislead the public and amplify extremist views because the labels do not reflect publicly known harmful roles. Google says the subtitles are generated automatically and that custom subtitles cannot be accepted. No response to the findings is reported.
- Company involved
- Google
- AI system involved
- Google Autocomplete
10 source articles · read the reporting →
Broadcasters sue Naver for using news content in AI training without permission
KBS, MBC, and SBS filed a copyright lawsuit against Naver, alleging that Naver used their news articles to train its generative AI platforms HyperCLOVA and HyperCLOVA X without permission. The broadcasters' association had previously sent notices prohibiting such use and requested compensation, but Naver did not respond. Naver has since revised its terms and refrained from using news data without consent.
- Company involved
- Naver
- AI system involved
- HyperCLOVA, HyperCLOVA X
5 source articles · read the reporting →
Good Daily operates AI-generated local news network without disclosure
Good Daily Inc., run by Matthew Henderson, operates hundreds of AI-generated local newsletters across the U.S. that aggregate and summarize news from local outlets. The newsletters use large language models without disclosing this to subscribers, and feature fabricated testimonials. Some readers reported being signed up without their consent. Local news outlets have criticized the network for not driving meaningful traffic and for undermining trust.
- Company involved
- Good Daily Inc.
- AI system involved
- Good Daily
3 source articles · read the reporting →
AI-driven Shadow Play network spreads pro-China propaganda on YouTube
The Australian Strategic Policy Institute (ASPI) revealed an extensive network of at least 30 YouTube channels, called Operation Shadow Play, that used generative AI to produce and publish pro-China and anti-US content. The network exploited YouTube's algorithmic recommendation system to cross-promote videos, gaining about 730,000 subscribers and 120 million views. The operation is alleged to be a coordinated influence campaign, possibly directed by a Mandarin-speaking controller, though the specific actor is unknown.
- Company involved
- Not specified (the operator is unknown)
- AI system involved
- Shadow Play
4 source articles · read the reporting →
RT launches AI-generated presenters without consistent disclosure
RT unveiled AI-generated 'journalists' Anastasia and Alexander on its Spanish-language channel RT Español. In 70% of the videos reviewed, the broadcasts failed to disclose that these presenters were artificial. The presenters covered a range of topics, including controversial and politically sensitive subjects, raising ethical concerns about transparency and potential manipulation of audiences.
- Company involved
- RT
- AI system involved
- Anastasia and Alexander (AI-generated presenters)
3 source articles · read the reporting →
Game Offline Lore used an AI clone of Mark Brown's voice in Doom videos
Mark Brown, creator of Game Maker's Toolkit, says YouTube channel Game Offline Lore used an AI-generated clone of his voice to narrate videos about Doom without his consent. He says the channel has removed comments flagging the videos and is likely earning ad revenue from them. Brown filed a privacy complaint with YouTube, which says it is reviewing the content. The videos remain live.
- Company involved
- Game Offline Lore
4 source articles · read the reporting →
Big Tech companies used YouTube videos to train AI without consent
Proof News found that subtitles from 173,536 YouTube videos were used by companies including Anthropic, Nvidia, Apple, and Salesforce to train AI models. The dataset, called YouTube Subtitles, was created by EleutherAI and published in 2020. Creators were not aware and some have expressed frustration, calling it theft. The companies have acknowledged using the dataset but argue it was publicly available.
- Company involved
- Anthropic, Nvidia, Apple, Salesforce, Bloomberg, Databricks
- AI system involved
- Claude, OpenELM
10 source articles · read the reporting →
Deepfake targeting France 24 journalist Catalina Marchant de Abreu ahead of Paris Olympics
A deepfake video impersonating France 24 journalist Catalina Marchant de Abreu was circulated on social media on July 6, 2024. The video falsely claimed she had investigated the safety of Paris' water system and reported dangerous viruses. The deepfake was attributed to Russian disinformation channels aiming to undermine the Paris Olympics. The video was identified and debunked by the account 'Random Osint'.
- Company involved
- France 24
5 source articles · read the reporting →
YouTube's algorithm serves disturbing fake cartoons to children
Thousands of videos on YouTube mimic popular children's cartoons like Peppa Pig and Frozen but contain violent, frightening or inappropriate content. The platform's recommendation system surfaces these videos to young children, who may not distinguish them from genuine cartoons. Parent Laura June reported her three-year-old daughter was exposed to a sadistic dentist scene in a fake Peppa Pig video. YouTube stated it removes flagged content and recommends parental controls, but experts argue the platform should provide better child protection.
- Company involved
- YouTube
- AI system involved
- YouTube recommendation algorithm
10 source articles · read the reporting →
YouTube's Recommendation Algorithm Found to Radicalise Users Towards Far-Right Content
A 2019 study audited YouTube's recommendation algorithm and found that it systematically suggests increasingly extreme content, leading users from milder right-wing channels to Alt-right and white supremacist content. The research analysed over 79 million comments and 2 million recommendations, showing a migration of users towards radicalisation. The findings support concerns that YouTube's algorithm creates a radicalisation pipeline. The study calls for greater transparency and public discussion on the role of algorithms.
- Company involved
- YouTube
- AI system involved
- YouTube recommendation algorithm
1 source article · read the reporting →
YouTube sued by black creators over discriminatory content moderation algorithms
A group of black YouTube creators filed a lawsuit against YouTube, alleging that its content moderation algorithms systematically remove their videos or limit advertising revenue without explanation. The suit, filed in federal court in Northern California, claims racial discrimination. The creators say the algorithms discriminate based on race, as a YouTube official allegedly acknowledged in 2017. The case is pending.
- Company involved
- YouTube
1 source article · read the reporting →
YouTube applies machine learning post-processing to Shorts without opt-out
YouTube announced an experiment on select Shorts using traditional machine learning to unblur, denoise, and improve clarity during processing. Creators reported that this altered their videos' appearance without an opt-out option, and some accused YouTube of misleading about the use of AI. YouTube acknowledged the experiment and said it would consider feedback, but did not offer a rollback or opt-out.
- Company involved
- YouTube
- AI system involved
- YouTube Shorts post-processing
7 source articles · read the reporting →
SEC charges YouPlus and CEO with defrauding investors
The SEC charged machine-learning startup YouPlus and its CEO Shaukat Shamim with defrauding investors. Shamim allegedly made false statements about the company's revenue and customer numbers, including providing falsified bank statements. The scheme unravelled when Shamim confessed to investors that the company had earned less than $500,000 and had only four paying customers since 2013. The SEC is seeking permanent injunctions, civil penalties, and an officer-and-director bar.
- Company involved
- YouPlus
- AI system involved
- YouPlus machine-learning tool
1 source article · read the reporting →
Beverly Hills Police Play Copyrighted Music to Trigger AI Copyright Filters on Activists' Livestreams
Police in Beverly Hills, California, played copyrighted music during interactions with activist Sennett Devermont, who was livestreaming. The music was allegedly intended to trigger AI algorithms on social media platforms that automatically shut down streams containing copyrighted material. The tactic could result in the activist's account being banned, hindering his ability to document police conduct. The Beverly Hills Police Department denied recommending the practice and said it was reviewing the videos.
1 source article · read the reporting →
YouTube's Algorithm Radicalised Caleb Cain, Leading to Far-Right Extremism
Caleb Cain, a liberal college dropout, turned to YouTube for self-help and was recommended videos by far-right creators like Stefan Molyneux. Over nearly five years, the platform's recommendation algorithm fed him increasingly extreme content, leading to his radicalisation. He eventually renounced the alt-right and spoke out about his experience, while YouTube announced policy changes to limit extremist content.
- Company involved
- YouTube
- AI system involved
- YouTube recommendation algorithm
2 source articles · read the reporting →
YouTube Algorithm Facilitates Child Sexual Exploitation Network, Monetised by Brands
In February 2019, YouTuber MattsWhatItIs revealed that YouTube's recommendation algorithm was connecting users to a network of videos containing child sexual exploitation content. The algorithm enabled paedophiles to trade contact information and links to child pornography in comments. The videos were being monetised by YouTube and major brands, including McDonald's and Disney. YouTube responded by flagging videos and disabling comments, but the content remained accessible and monetised.
- Company involved
- YouTube
- AI system involved
- YouTube recommendation algorithm
1 source article · read the reporting →
YouTube sued by LGBT creators over alleged algorithmic discrimination
A class action lawsuit filed by 12 LGBT YouTubers alleges that YouTube's algorithm discriminates against their content by demonetizing, age-restricting, and limiting the visibility of videos. The plaintiffs claim the automated system makes decisions within moments of upload, without human review, and that YouTube has failed to address the bias. The lawsuit, which includes a recording of a YouTube manager citing 'the gay thing' as a reason for restricting content, seeks systemic change and transparency.
- Company involved
- YouTube
1 source article · read the reporting →
X's algorithm amplifies right-wing and extreme content, Sky News investigation finds
A Sky News investigation found that X's algorithm amplifies right-wing and extreme content. The Data and Forensics team conducted a nine-month investigation into the platform's recommendation system. The findings indicate that the algorithm boosts such content, potentially influencing political discourse in the UK.
- Company involved
- X
- AI system involved
- X's algorithm
4 source articles · read the reporting →
Facebook exempted Trump and other high-profile users from content enforcement rules.
An internal Facebook program called XCheck granted special treatment to millions of high-profile users, including former President Donald Trump and Senator Elizabeth Warren, exempting them from content enforcement rules. The program, initially for quality control, was expanded to whitelist users deemed newsworthy or PR-risky, with the system failing to enforce rules against them. Internal documents revealed that the program made mistakes, often wrongly taking action against high-profile users. Facebook has attempted to phase out the practice but has struggled.
- Company involved
- Facebook
- AI system involved
- XCheck
10 source articles · read the reporting →
YouTube's AI captions insert explicit language in children's videos
YouTube's automated captioning system has been found to insert explicit language into videos intended for children. A study of over 7,000 videos from top kids' channels found that 40% contained taboo words in captions, with 1% including highly inappropriate terms like 'bitch' and 'rape'. The AI mishears words such as 'brave' as 'rape' and 'corn' as 'porn'. YouTube acknowledged the issue and said it is working to improve the system.
- Company involved
- YouTube
- AI system involved
- YouTube automated captions
7 source articles · read the reporting →
Consumer groups ask FTC to investigate YouTube Kids app for deceptive ads
Several consumer groups filed a complaint with the Federal Trade Commission alleging that Google's YouTube Kids app mixes advertising with programming in ways that are deceptive to children. The complaint argues that the app uses host selling and undisclosed endorsements, which would be illegal on television. YouTube responded that it worked with child advocacy groups and disagreed with the contentions. The FTC has not yet acted on the complaint.
- Company involved
- Google
- AI system involved
- YouTube Kids
10 source articles · read the reporting →
True Crime Case Files YouTube channel exposed for AI-generated fake true crime stories
The YouTube channel True Crime Case Files, which launched in late 2023, gained millions of views with videos purporting to be true crime documentaries. An investigation by the Denver Post revealed that the stories, including one about a murder in Littleton, Colorado, were entirely fabricated and likely generated by AI. The channel did not disclose that the cases were not real, leading viewers to believe they were genuine. Experts warned that such AI-generated misinformation is a growing problem, with the channel potentially earning tens of thousands of dollars from the deceptive content.
- Company involved
- True Crime Case Files
- AI system involved
- True Crime Case Files
5 source articles · read the reporting →
YouTube's algorithm systematically exposes children to disturbing automated content
YouTube's recommendation algorithm and automated content generation systems have been producing and promoting disturbing videos targeting children. These videos, often generated by bots using keyword stuffing and stock animations, include frightening and abusive content. The article alleges that the platform's automated systems, without human oversight, systematically traumatise children at scale.
- Company involved
- YouTube
- AI system involved
- YouTube recommendation algorithm
10 source articles · read the reporting →
YouTube recommended AI-generated disinformation videos to children as educational content
BBC journalists found over 50 YouTube channels using generative AI to create false scientific videos disguised as educational content. YouTube's recommendation algorithm suggested these videos to children alongside legitimate content. Children who watched the videos believed the false information, such as alien conspiracies and electric pyramids. YouTube said it was committed to removing misinformation but did not comment on advertising revenue from the misleading videos.
- Company involved
- YouTube
3 source articles · read the reporting →
