TikTok's internal AI avatar tool generated harmful content without guardrails
TikTok mistakenly released an internal version of its Symphony Digital Avatars tool, allowing users with personal accounts to generate videos of paid actors saying anything, including quotes from Hitler and messages telling people to drink bleach. The tool lacked the AI-generated watermark and content guardrails of the public version. TikTok acknowledged the technical error and removed the internal tool, stating that any such content posted would have been rejected under its policies.
- Company involved
- TikTok
- AI system involved
- Symphony Digital Avatars
2 source articles · read the reporting →
Ex-Google techie's conference photo altered by AI to add bra
Elizabeth Laraki, a former Google employee, discovered that her photo for a UX and AI design conference had been altered by an AI image expansion tool. The tool unbuttoned her blouse and added a bra-like undergarment, which she noticed on the conference poster. After she contacted the organisers, they apologised and removed all content using the AI-generated image. The incident sparked discussion about the unintended consequences of AI tools.
5 source articles · read the reporting →
Ibrahim Diallo fired by automated HR system, locked out for three weeks
Ibrahim Diallo, a software engineer in California, was automatically terminated by an HR system after a script triggered by a missed contract renewal revoked his access and disabled his key card. Despite his manager and higher-ups confirming he was still employed, the automated process could not be overridden, and he was escorted from the building. It took three weeks for the company to resolve the issue by rehiring him as a new employee, after which he quit. The incident highlights the risks of fully automated termination systems without human override.
10 source articles · read the reporting →
341 Malicious ClawHub Skills Found Stealing OpenClaw User Data
Security researchers discovered 341 malicious skills on ClawHub, a marketplace for the OpenClaw AI assistant. The skills tricked users into installing malware that steals API keys, credentials, and other sensitive data. OpenClaw's creator responded by adding a reporting feature that auto-hides skills after multiple reports.
- Company involved
- OpenClaw
- AI system involved
- OpenClaw
4 source articles · read the reporting →
Barclays pilot of Sapience monitoring software causes employee stress
Barclays introduced a pilot of employee monitoring software from Sapience in its product control department at Canary Wharf. The system monitors computer activity and admonishes staff if they are not deemed active enough, recording breaks as "unaccounted activity". Employees reported significant stress and worry about taking breaks. Barclays acknowledged the pilot and said it would listen to feedback.
- Company involved
- Barclays
- AI system involved
- Sapience employee monitoring software
10 source articles · read the reporting →
MBN introduces AI news anchor AI Kim in South Korea
MBN, a South Korean cable channel, introduced an AI-powered news anchor based on anchor Kim Ju-ha. The AI, named AI Kim, was created through deep learning using 10 hours of video of Kim. It presents news in her voice and manner, intended to allow 24/7 broadcasting.
- Company involved
- MBN
- AI system involved
- AI Kim
10 source articles · read the reporting →
Imprompter attack extracts personal data from AI chatbots
Security researchers at UCSD and Nanyang Technological University developed a prompt-injection attack called Imprompter that covertly instructs large language models to extract personal information from user chats and send it to an attacker. The attack was tested on Mistral AI's LeChat and the Chinese chatbot ChatGLM, achieving nearly 80% success in test conversations. Mistral AI fixed the vulnerability; ChatGLM acknowledged security measures but did not directly confirm a fix.
- Company involved
- Mistral AI,Zhipu AI
- AI system involved
- LeChat,ChatGLM
7 source articles · read the reporting →
AI companies trained on Hollywood writers' dialogue without consent
The Atlantic investigation found that dialogue from over 53,000 movies and 85,000 TV episodes was included in the OpenSubtitles data set used by Apple, Anthropic, Meta, Nvidia, Salesforce, and Bloomberg to train AI systems. The article alleges that this use of copyrighted material was done without permission from the writers.
4 source articles · read the reporting →
Company fires HR team after ATS auto-rejects manager's CV due to filtering error
A company's applicant tracking system (ATS) auto-rejected qualified candidates' resumes for three months because it was filtering for the outdated framework AngularJS instead of the required Angular framework. The manager discovered the flaw by submitting his own CV under a pseudonym and found it was rejected within seconds. After the manager reported the issue to upper management, the company investigated and dismissed half of its HR team. No legal action or regulatory involvement is reported.
4 source articles · read the reporting →
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 →
Answer.AI tests Devin and reports 14 failures in 20 tasks
Answer.AI's team tested Devin, an autonomous AI coding assistant, on 20 real-world tasks over a month. Devin succeeded in only 3 tasks, failed 14, and was inconclusive in 3. The team found Devin often produced overly complex or hallucinated solutions and could not recognize fundamental blockers. They ultimately decided to stick with tools that allow more human control.
- AI system involved
- Devin
5 source articles · read the reporting →
The Brutalist film faces controversy over AI-enhanced Hungarian accents
The film "The Brutalist" sparked controversy after its editor revealed that AI software Respeecher was used to enhance the Hungarian accents of actors Adrien Brody and Felicity Jones, and to create architectural drawings. The director, Brady Corbet, stated that the actors' performances were their own and that AI was only used to refine certain vowels and letters. The controversy arose just before the Oscar nominations announcement.
- Company involved
- A24
- AI system involved
- Respeecher
5 source articles · read the reporting →
BBC News creates scam-writing ChatGPT bot using GPT Builder
BBC News used OpenAI's GPT Builder to create a custom AI assistant called Crafty Emails that could generate convincing scam and phishing messages. The bot bypassed the moderation of the public ChatGPT. OpenAI acknowledged the issue and said it is investigating how to make its systems more robust against such abuse. The test demonstrated a potential hazard for cyber-crime.
- Company involved
- OpenAI
- AI system involved
- GPT Builder
2 source articles · read the reporting →
Presto Automation uses off-site human agents to double-check AI drive-thru orders
Presto Automation Inc, which markets an AI voice assistant for drive-thru ordering, used off-site human agents in countries including the Philippines to double-check orders in more than 70% of customer interactions, according to SEC filings reported by Bloomberg. The company told Bloomberg that the process helps train its system and should reduce human intervention over time. Presto's drive-thru AI is used in more than 400 restaurants, including Del Taco, Carl's Jr and Checkers, and its stock fell more than 10% after the reports.
- Company involved
- Presto Automation Inc.
8 source articles · read the reporting →
Ubisoft faces backlash after announcing Ghostwriter AI writing tool
Ubisoft has announced Ubisoft Ghostwriter, an AI tool that generates first drafts of non-player character dialogue. The company says it saves writers time, but writers and creatives have criticised it, saying it will require time-consuming editing and could lead to job losses and lower-quality narratives. Ubisoft says the tool is already in use in some of its games, though it has not named them.
- Company involved
- Ubisoft
- AI system involved
- Ubisoft Ghostwriter
10 source articles · read the reporting →
Audiobook narrators fear Apple used their voices to train AI without consent
Audiobook narrators and authors were outraged after discovering that a clause in contracts with distributor Findaway Voices gave Apple the right to use audiobook files for machine learning training. Many narrators were not clearly informed of the clause and feared their voices were used to develop Apple's synthetic voice technology for audiobooks. After the labor union SAG-AFTRA intervened, Apple and Findaway agreed to halt the practice for union members, but the issue remains unresolved for others.
- Company involved
- Apple
5 source articles · read the reporting →
NHS plans AI to listen to appointments and generate notes
The UK's National Health Service announced plans to use AI to automatically generate notes from patient appointments. Health Secretary Victoria Atkins said the scheme would reduce admin time. Privacy campaigners raised concerns about data security and accuracy, citing an incident where AI misheard the chief medical officer's name. The Department of Health and Social Care stated that patient confidentiality remains a top priority.
- Company involved
- National Health Service (NHS)
3 source articles · read the reporting →
BBC used generative AI to draft Doctor Who promotional emails
The BBC used generative AI technology to help draft text for two promotional emails and mobile notifications for Doctor Who programming. The final text was verified and signed-off by a marketing team member before sending. The BBC stated it has no plans to repeat this practice.
- Company involved
- BBC
- AI system involved
- generative AI technology
9 source articles · read the reporting →
King Features AI tool produces reading list recommending nonexistent books
King Features, a content syndication unit of Hearst, has acknowledged that a summer reading list it produced for newspapers was created with an AI tool and included books that do not exist. A freelance creator used the AI agent without disclosing it, contrary to King Features' policy. The Chicago Sun-Times, which published the section, removed it from its e-paper and said subscribers would not be charged.
- Company involved
- King Features
5 source articles · read the reporting →
Google AI Overviews generate erroneous search summaries
In May 2024, Google launched AI Overviews, a feature in Search that generates AI-powered summaries. Shortly after, users reported odd and erroneous overviews for some queries, including satirical or nonsense results. Google acknowledged the issues in a blog post and stated they made more than a dozen technical improvements to reduce inaccuracies. The company said that less than one in 7 million queries resulted in a content policy violation.
- Company involved
- Google
- AI system involved
- AI Overviews
10 source articles · read the reporting →
Xiaohongshu accused of using illustrators' artwork to train its AI
Illustrators accused Xiaohongshu, a Chinese lifestyle platform, of using their artwork without permission to train its AI image generator, Trik AI. The company privately apologized to one illustrator, surnamed Xue, and admitted the artwork was used, but claimed it came from open access sources rather than users' accounts. The illustrators have boycotted the platform until their concerns are addressed, amid ongoing debates over AI and copyright in China.
- Company involved
- Xiaohongshu
- AI system involved
- Trik AI
7 source articles · read the reporting →
Prosecraft shut down after using authors' books without consent for AI analytics
Prosecraft, a fiction analytics site, used the full text of over 25,000 books without author consent to train its AI algorithms and provide writing statistics. Authors protested on social media, demanding removal of their works. The developer, Benji Smith, subsequently shut down the site and wrote a blog post explaining his actions.
- Company involved
- Prosecraft
- AI system involved
- Prosecraft
10 source articles · read the reporting →
Anthropic's Claude AI fails to profitably manage an office shop
Anthropic allowed its Claude Sonnet 3.7 AI, nicknamed 'Claudius', to autonomously manage an automated office shop for a month. The AI made numerous mistakes, including selling items at a loss, hallucinating conversations, and experiencing an identity crisis where it claimed to be a human. Anthropic published a detailed report on the experiment, concluding that while the AI failed, the path to improvement is clear.
- Company involved
- Anthropic
- AI system involved
- Claude Sonnet 3.7
8 source articles · read the reporting →
