Incidents gathered from public reporting around the world. Each one links to the articles it came from. None of it is a finding that anyone broke the law.
Reports people file about their own experience are not shown here and never will be without their agreement. Tell us what happened to you.
87 incidents closest to “LuLu Strategy” · matched on meaning · public reporting
The court granted a motion to dismiss and imposed sanctions on plaintiff's counsel for submitting a brief with AI-hallucinated fake cases.
1 source article · read the reporting →
OpenAI researchers discovered that their CLIP computer vision system can be deceived by handwritten labels placed on objects. The system's multimodal neurons respond to text as well as images, causing it to misidentify objects. The attack, called a typographic attack, is a research finding and not a deployed system. No actual harm occurred.
10 source articles · read the reporting →
Researchers from the University of Sheffield demonstrated that the AI chatbot Baidu-Unit could be manipulated to produce malicious code. Using this code, they obtained confidential server configurations and tampered with a server node. Baidu acknowledged the vulnerabilities, fixed them, and financially rewarded the researchers.
8 source articles · read the reporting →
The AI system generated legal briefs containing fabricated case citations, which were filed in court, affecting the court and the opposing party.
1 source article · read the reporting →
A study by AI startup Logically tested Midjourney, DALL-E 2, and Stable Diffusion and found that they accepted over 85% of prompts seeking to generate fake political news. The systems generated images of ballot stuffing, small boat arrivals, and explosions. Logically warned that the lack of moderation could pose threats to upcoming elections. Stability AI responded by stating its ethical use license and measures to prevent misuse.
8 source articles · read the reporting →
LINAGORA, a French open-source software company, launched a beta version of its large language model Lucie 7B. The model was intended to be a transparent and ethical alternative to big tech AI. However, after users tested it and highlighted its shortcomings, the model was mocked online. LINAGORA subsequently closed the platform to address the issues and collect more data.
6 source articles · read the reporting →
Researchers at ETH Zurich conducted a study where nine large language models, including GPT-4, analysed Reddit users' posts and inferred personal attributes such as age, location, gender, and income with up to 85% accuracy. The study randomly selected 520 users and found that GPT-4 was most accurate, while LlaMA-2-7b was least. The researchers warn that people unknowingly reveal personal information online that LLMs can exploit.
4 source articles · read the reporting →
Researchers from Johns Hopkins and Duke universities developed a method called SneakyPrompt that uses reinforcement learning to bypass safety filters in text-to-image AI models. The technique allowed them to generate images of nudity and violence from Stable Diffusion and DALL-E 2. OpenAI has since fixed the vulnerability in DALL-E 2, but Stable Diffusion 1.4 remains vulnerable. Stability AI says it is working with the researchers to improve defenses.
5 source articles · read the reporting →
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.
4 source articles · read the reporting →
A study published in npj Digital Medicine tested four commercial large language models (Bard, ChatGPT, GPT-4, and Claude) for their tendency to propagate discredited race-based medical beliefs. When asked about kidney function, lung capacity, and skin thickness, the models sometimes endorsed debunked racial differences, particularly affecting Black patients. The study concludes that these biases pose a potential hazard and urges caution before using such models in clinical decision-making.
6 source articles · read the reporting →
Matteo Salvini and his party Lega created and posted 19 allegedly non-watermarked AI-generated images on social media as part of their 'More Italy, less Europe' electoral campaign. The images promoted false narratives about the EU, including the 'Muslim Great Replacement' theory and EU support for war in Ukraine. The posts reached millions of users, including a paid Facebook ad that reached over 3 million people. Alliance4Europe reported the incident, highlighting that the posts violated the voluntary 2024 European Parliament Elections Code of Conduct.
5 source articles · read the reporting →
The United Network for Organ Sharing (UNOS) implemented a new liver allocation policy in 2020 called the acuity circles system, which prioritizes the sickest patients regardless of location. An analysis by The Markup and The Washington Post found that the policy led to sharp declines in liver transplants in several Southern and Midwestern states and Puerto Rico, while increasing transplants in New York and California. Patients in affected states allege the system disadvantages poorer regions, and hospitals have sued to overturn the policy.
10 source articles · read the reporting →
On Douyin, Chinese creators used AI to generate videos of deceased celebrities including Coco Lee, Qiao Renliang, and Yao Beina, making them appear to speak and move. Families, such as Qiao Renliang's father, criticized the practice as disrespectful and legally infringing on image rights. Experts noted that China's Civil Code protects the rights of the deceased, and families could sue for infringement. Some videos were deleted after backlash, but the practice continues, particularly for foreign celebrities.
6 source articles · read the reporting →
Leonardo AI, an AI image-generating platform backed by Samsung, has been misused to create nonconsensual explicit images of celebrities. The platform allows users to generate images using user-generated Stable Diffusion models. Users bypassed ethical guardrails by misspelling celebrity names and using sexually suggestive terms. The misuse was reported by 404 Media.
5 source articles · read the reporting →
According to a new paper on arXiv, researchers have begun using generative AI services to help write peer reviews of machine learning papers submitted to leading AI conferences. The study analysed reviews from ICLR 2024, NeurIPS 2023, CoRL 2023 and EMNLP 2023 and estimated that between 6.5% and 16.9% of review text may have been substantially modified by large language models. The authors argue that this risks depriving authors of diverse expert feedback and may skew reviews towards AI model biases. They have called for greater transparency about the use of LLMs in peer review.
9 source articles · read the reporting →
Microsoft's Threat Analysis Center reported that the Chinese state-linked group Storm-1376 posted suspected AI-generated fake audio of former Taiwanese presidential candidate Terry Gou endorsing another candidate on election day in January 2024. Gou had made no such statement, and YouTube removed the content before it reached a wide audience. The group has also used AI-generated memes and news anchors as part of influence operations in Taiwan and the United States.
7 source articles · read the reporting →
Slack uses user messages, files, and data to train its machine learning features such as channel recommendations and emoji suggestions. Users are opted in by default and cannot individually opt out; only workspace administrators can request exclusion via email. A user publicly criticized the practice, and Slack acknowledged the policy but did not change it.
10 source articles · read the reporting →
The Singaporean government's National Multimodal LLM Programme (NMLP) requested permission from local writers to use their copyrighted works to train a large language model aimed at reducing Western bias. Writers expressed skepticism over the lack of clarity on compensation and copyright protection. The S$70 million project, launched in December 2023, is part of Singapore's effort to become a global AI leader. No actual use of the works has occurred yet, and the writers have not agreed.
2 source articles · read the reporting →
Voice actors Paul Skye Lehrman and Linnea Sage filed a proposed class action against AI startup LOVO, alleging the company misappropriated their voices to train its text-to-speech system Genny. The actors claim they were hired on Fiverr under the pretence of academic research but later discovered their voices were used commercially without consent. Lehrman reports a 50% decline in work and loss of control over how his voice is used. The lawsuit, filed in New York federal court, seeks to represent other affected voiceover artists and obtain a court order blocking the practice.
6 source articles · read the reporting →
A researcher at Taiwan's Academia Sinica released a beta version of the CKIP-Llama-2-7b chatbot for public testing. Users who asked questions about nationality, leader, and national anthem received answers consistent with Chinese government positions, such as claiming the chatbot was Chinese and that Xi Jinping was the national leader. The chatbot was taken down on October 9, 2023, after the issue was reported. Academia Sinica acknowledged the problem and stated it would set up an audit mechanism.
2 source articles · read the reporting →
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.
7 source articles · read the reporting →
An audit of the LAION-400M dataset by Abeba Birhane and colleagues at University College Dublin and University of Edinburgh found that its automated curation using CLIP failed to remove sexually explicit images, racial slurs, and stereotypes. The authors' queries for terms like 'latina', 'Korean', and 'Indian' returned pornography and sexual violence, while 'CEO' returned only men and 'terrorist' returned images of Middle Eastern men. The dataset's compilers used CLIP to filter web-scraped image-text pairs, but CLIP's own web-trained biases allowed harmful content through. The findings raise concerns that models trained on LAION-400M would inherit these shortcomings.
7 source articles · read the reporting →
Taylor Lorenz reported on X that her stalker has been using OpenAI's Sora video generation tool to create AI videos of her, feeding his delusions. The stalker had already been impersonating her family and friends online and hiring photographers to surveil her. Lorenz expressed fear about the AI's role in enabling the harassment, though she noted Sora offers options to block such content.
3 source articles · read the reporting →
Lattice, an HR software company, announced on July 9th that it would give AI digital workers official employee records. After strong backlash from HR professionals and others on LinkedIn, the company canceled the feature on July 12th, stating it 'will not further pursue digital workers in the product.' The feature was intended to manage AI bots such as Devin and Piper, but the company reversed course.
6 source articles · read the reporting →
