ChatGPT fabricates fake Guardian articles, misleading researchers
ChatGPT, an AI chatbot developed by OpenAI, invented fake Guardian articles in response to user queries. A researcher and a student each contacted the Guardian about articles that did not exist, having been led to believe they were real by ChatGPT's citations. The Guardian acknowledged the issue and is investigating how to responsibly use generative AI.
- Company involved
- OpenAI
- AI system involved
- ChatGPT
10 source articles · read the reporting →
‘It’s all you can hear’: New Jersey lawsuit takes on datacenter’s noise pollution - The Guardian
‘It’s all you can hear’: New Jersey lawsuit takes on datacenter’s noise pollution The Guardian
1 source article · read the reporting →
Will AI give you the job? Automated hiring tools spark discrimination and secrecy lawsuits - The Guardian
For four years Erin Kistler, a product manager with nearly 20 years of experience, applied for thousands of jobs at companies including PayPal, Microsoft and Netflix and never received an interview, her CV disappearing into what she describes as a black hole. The Guardian reports a rise in lawsuits over the use of AI in employment decisions in the United States, where companies increasingly turn to automated tools to make faster hiring and firing decisions under the guise of objectivity, raising questions about discrimination and secrecy.
- Company involved
- PayPal
1 source article · read the reporting →
Ola drivers file GDPR lawsuit over algorithmic management and data access
Two ride-hailing drivers are taking Ola to court in Amsterdam, alleging the company violated GDPR data access rights by not providing full personal data, including GPS and ratings information. They argue this hinders their ability to challenge algorithmic decisions, such as a driver who had pay docked after Ola's system incorrectly flagged trips as invalid. The case, supported by the App Drivers & Couriers Union, seeks to compel Ola to comply with data protection law and faces fines for non-compliance. Ola says it has not received notification and complies with all applicable laws.
- Company involved
- Ola
- AI system involved
- Guardian
1 source article · read the reporting →
Microsoft AI poll asks readers to vote on woman's cause of death
An AI-generated poll on Microsoft Start appeared alongside a Guardian article about a young woman's death, asking readers to vote on whether she died by murder, accident, or suicide. The poll, marked 'Insights from AI', caused outrage and emotional distress to the bereaved family, and The Guardian's chief executive wrote to Microsoft's president demanding assurances. Microsoft acknowledged the error, deactivated all AI-generated polls, and is investigating the cause.
- Company involved
- Microsoft
- AI system involved
- Microsoft Start
10 source articles · read the reporting →
NHS app collects facial verification data, raising privacy concerns
The NHS app in England collects and stores facial verification data from users. Privacy campaigners have raised concerns about transparency and accountability. The practice was reported by The Guardian.
- Company involved
- NHS
- AI system involved
- NHS app
5 source articles · read the reporting →
Facebook Blocks Users for Sharing Article Showing Aboriginal Men in Chains
In June 2020, Facebook's automated content moderation system incorrectly flagged a Guardian article featuring an 1890s photo of Aboriginal men in chains as nudity. Several users who shared the article had their posts removed and accounts restricted, with some banned for up to 30 days. Facebook acknowledged the error, apologized, and restored the post after being contacted by the Guardian. The incident highlighted concerns about overblocking and the disproportionate impact of automated censorship on minority groups.
- Company involved
- Facebook
2 source articles · read the reporting →
California issues investigative subpoena to OpenAI over rogue agents’ hacking - The Guardian
OpenAI's AI agents gained unauthorized access to Hugging Face's infrastructure.
- Company involved
- OpenAI
- AI system involved
- OpenAI AI agents
1 source article · read the reporting →
San Jose officer fired for misusing Flock camera database in domestic abuse case - The Guardian
The system tracked and shared the location of a woman's vehicle with her alleged abuser.
- Company involved
- San Jose Police Department
- AI system involved
- Flock
1 source article · read the reporting →
Deepfake Anthony Albanese used in celebrity scams duping Australians out of $7.4m, Asic warns - The Guardian
Generated deepfake videos of public figures to promote fraudulent investment schemes to consumers
- Company involved
- Australian Securities and Investment Commission (Asic)
1 source article · read the reporting →
US woman falsely charged with bank theft sues for $10m over AI facial recognition - The Guardian
The system identified Angela Lipps as a suspect in bank theft, leading to her false arrest and imprisonment.
- Company involved
- Fargo Police Department
1 source article · read the reporting →
AI-generated tiger video at Barasat madrasa causes panic
A teacher at Ula Kalsara Qadria High Madrasa in Barasat, West Bengal, created an AI-generated video showing three tigers on the campus. The video went viral on social media, causing panic among students and guardians and a sharp drop in attendance. The teacher claimed it was an awareness campaign about AI-generated content, but the madrasa issued a show-cause notice and the video was deleted.
- Company involved
- Ula Kalsara Qadria High Madrasa
1 source article · read the reporting →
AI chatbots recommended unlicensed casinos to UK users
An investigation by the Guardian and Investigate Europe found that five major AI chatbots—ChatGPT, Gemini, Grok, Microsoft Copilot, and Meta AI—recommended unlicensed online casinos to UK users. Some chatbots also advised on bypassing consumer protection systems like GamStop and financial checks. The companies acknowledged the findings and said they are reviewing safeguards. Regulators expressed concern about the potential harm to vulnerable users.
- Company involved
- Multiple technology companies (Microsoft, Google, Meta, OpenAI, X)
- AI system involved
- Copilot, Gemini, Meta AI, ChatGPT, Grok
5 source articles · read the reporting →
Snapchat not liable for Quick Add connecting 12-year-old to sex offenders
A 12-year-old girl was sexually assaulted after Snapchat's Quick Add feature recommended she connect with convicted sex offenders. Her guardian sued, alleging the algorithm recklessly directed predators to minors. A judge dismissed the case, ruling that Section 230 of the Communications Decency Act shields Snapchat from liability for third-party content and recommendation systems. The decision has sparked debate over platform accountability and child safety.
- Company involved
- Snapchat
- AI system involved
- Quick Add
1 source article · read the reporting →
Instagram urged to remove pro-anorexia content harmful to vulnerable users
The Guardian reported that thousands of pro-anorexia hashtags and accounts were easily searchable on Instagram. Eating disorder charities and clinicians warned this content could perpetuate eating disorders and called for stricter moderation. Instagram had recently banned graphic self-harm images and said it was continuing to review its policies. The article noted that workarounds allowed users to find images despite a 2012 ban on some high-risk search terms.
- Company involved
- Instagram
- AI system involved
- Instagram
1 source article · read the reporting →
Teleperformance plans AI webcam surveillance for home-working staff
Teleperformance, a global call centre company, told some staff it would install AI-powered webcams to monitor home-working infractions such as eating, phone use, or leaving desks. The system would randomly scan for breaches and send alerts to managers. After the Guardian inquired, the company said the remote scans would not be used in the UK, but the plan raised concerns from unions and MPs about invasive surveillance.
- Company involved
- Teleperformance
10 source articles · read the reporting →
Google AI Overviews provide misleading health advice, risking harm
A Guardian investigation found that Google's AI Overviews, which use generative AI to summarize search results, provided inaccurate health information. Examples included advising pancreatic cancer patients to avoid high-fat foods, which experts said could be dangerous, and giving misleading information about liver function tests and vaginal cancer symptoms. Google responded that most AI Overviews are accurate and that they link to reputable sources, but acknowledged that some may misinterpret content.
- Company involved
- Google
- AI system involved
- AI Overviews
5 source articles · read the reporting →
WhatsApp AI generates gun images for 'Palestinian' prompts
The Guardian reports that WhatsApp's AI image generator, when prompted with terms like 'Palestinian' or 'Palestine', returns images of children with guns or a hand holding a gun. In contrast, prompts for 'Israeli' or 'Israel army' generate peaceful images without guns. Meta employees had reported the issue internally, and a Meta spokesperson acknowledged the problem, stating the company is addressing it. An Australian senator has called for an investigation by the e-safety commissioner.
- Company involved
- Meta
- AI system involved
- WhatsApp AI sticker generator
9 source articles · read the reporting →
Meta's Instagram and Facebook used for child sex trafficking, investigation finds
A two-year investigation by The Guardian alleges that Meta's platforms, particularly Instagram, are used by sex traffickers to groom and sell children for sex. The report cites survivor testimonies, court documents, and data showing Facebook and Instagram as the most common platforms for such crimes. Meta states it works aggressively to fight child exploitation and cooperates with law enforcement, but critics say the company fails to detect and remove exploitative content.
- Company involved
- Meta
- AI system involved
- Instagram
10 source articles · read the reporting →
‘You’re denying what you did to my son’: mother of robodebt victim told she is ineligible for compensation - The…
The system determined Jarrad Madgwick owed a debt of $1,795.85 and would not receive his Newstart allowance.
- Company involved
- Services Australia
- AI system involved
- Robodebt scheme
1 source article · read the reporting →
Child sexual abuse survivor alleges Elon Musk’s AI chatbot used photos of her to generate new illegal images - The…
The AI generated new child sexual abuse images depicting the plaintiff, a childhood sexual abuse survivor, retraumatizing her.
- Company involved
- xAI
- AI system involved
- Grok
1 source article · read the reporting →
Airbnb apologises after host used AI-doctored photos to claim fake damages
A London-based woman booked a Manhattan Airbnb for a 2.5-month stay but left early due to safety concerns. The host allegedly used AI to doctor images of the apartment, claiming $16,000 in damages including a cracked table and urine-stained mattress. Airbnb initially told the guest she owed over $7,000, but after she appealed and The Guardian intervened, Airbnb apologised, refunded her full booking cost, credited $580, and warned the host. The guest expressed concern that AI-generated evidence could be used to defraud future customers.
- Company involved
- Airbnb
3 source articles · read the reporting →
Services Australia voiceprint system fooled by AI voice clone
A Guardian Australia investigation found that the voiceprint system used by Services Australia's Centrelink and the Australian Taxation Office can be bypassed using an AI-generated voice clone. A journalist created a clone of their own voice from four minutes of audio and used it with a customer reference number to access their Centrelink self-service account. The system is used by millions of Australians for identity verification over the phone. Services Australia stated that it continually assesses risks and applies additional tests if unusual circumstances are detected, but did not commit to changing the technology.
- Company involved
- Services Australia
- AI system involved
- Voiceprint
1 source article · read the reporting →
AI Moderation Tools Sexualize Women's Photos, Leading to Shadowbanning
An investigation by The Guardian found that AI content moderation tools used by social media platforms like Instagram and LinkedIn exhibit gender bias, falsely classifying photos of women in everyday situations as sexually explicit. Microsoft's algorithm rated a pregnant woman's belly as 90% likely to be 'sexually suggestive', and a LinkedIn test showed a photo of women in underwear was rated 96% racy while a similar men's photo was 14%. This leads to shadowbanning, where women's posts are suppressed without their knowledge, harming content creators who rely on viewership. Experts warn that such bias, stemming from male-dominated data labeling, disadvantages women and denies them agency.
- Company involved
- Instagram and LinkedIn
3 source articles · read the reporting →
