Researchers Develop Adversarial Method to Detect AI-Generated Social Bot Content
A Swedish research team fabricated a fictional eye condition called 'bixonimania' and seeded it online to test whether AI chatbots could detect misinformation, finding that major systems including ChatGPT, Google Gemini, and Microsoft Copilot all presented the fake disease as real. The bogus condition was introduced via Medium blog posts and preprints laden with obvious red flags, including a lead author whose name translates to 'lying loser' and a paper that explicitly stated 'This entire paper is made up.' The experiment exposes a systemic vulnerability in how large language models ingest and reproduce unverified internet content, with implications for public health misinformation.
Almira Osmanovic Thunström and colleagues at the University of Gothenburg designed the bixonimania experiment in early 2024 to demonstrate how large language models (LLMs) process and redistribute misinformation. The team published two Medium blog posts and two preprints — since removed — describing the nonexistent condition, which they deliberately filled with absurd signals: funding attributed to the 'Galactic Triad' and Lord of the Rings, acknowledgments to the Starship Enterprise and fictional character Professor Ross Geller, and an explicit in-text declaration that the paper was fabricated. Despite these clues, by April 2024 Microsoft Copilot, Google Gemini, and ChatGPT were all describing bixonimania as a genuine medical condition, and the disease even acquired its own Wikipedia page. Compounding the problem, some human researchers cited the bogus preprints without apparently reading them. Misinformation experts quoted in coverage described the episode as a 'masterclass' in how false information propagates through interconnected digital and scientific systems. The findings underscore that LLMs trained on large web corpora inherit the misinformation present in that data, and that neither automated nor human review reliably caught content riddled with obvious fabrications.
What's missing
The article does not report whether the AI companies were notified or have since updated their models to correct the bixonimania responses.
What different sources said
- arXiv cs.CLCenter
SV-Detect: AI-generated Text Detection with Steering Vectors
- Smithsonian MagazineCenter
Scientists Invented a Disease to Test Whether A.I. Knew It Was Fake. Then, Chatbots Started Saying It Was Real
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.