Rabbit Hole

See what people are claiming online - rumors, conspiracy theories, and questionable content.

What is this?

We collect stories that are spreading online - things like conspiracy theories, fake videos, and propaganda - so you can see what's being said and make up your own mind about what's real.

Propaganda

Agenda-pushing content

Rumor

Unverified theories

Deepfake

AI-generated fakes

Influence Ops

Organized campaigns

Evidence = How much proof exists
Mystery = How intriguing it is
Rumor

Fake Argos promotions claim Samsung Galaxy phones sold at 99% discount

Facebook posts and adverts say the offer can be claimed by completing a short survey.

Evidence 4/5
Mystery 1/5
Jan 31, 2026 1
Rumor

What did Reform UK candidate Matt Goodwin say about Manchester in 2023?

Reform UK has accused Labour of sharing a misleading video featuring past comments from its Gorton and Denton by-election candidate.

Evidence 4/5
Mystery 2/5
Jan 31, 2026 1
Rumor

Currys is not offering a new Xbox for £3 as reward for completing a survey

Currys has said the offer being shared in Facebook posts and adverts is not a genuine one from them.

Evidence 4/5
Mystery 1/5
Jan 31, 2026 1
Rumor

Zara isn’t giving away a free box of clothes to people who fill out a survey

Facebook posts and adverts falsely claim the retailer is offering a free box of clothing and accessories to people who complete a questionnaire.

Evidence 4/5
Mystery 1/5
Jan 31, 2026 1
Rumor

IKEA doesn’t run ‘Mystery Box’ offer shown in Facebook posts

IKEA has confirmed this is not a genuine offer from the retailer.

Evidence 4/5
Mystery 1/5
Jan 31, 2026 1
Deepfake

Another AI-generated video of Donald Trump criticising Keir Starmer circulates online

The video appears to show Donald Trump telling Keir Starmer to concentrate on governing the UK. But it isn’t real.

Evidence 4/5
Mystery 2/5
Jan 31, 2026 2
Deepfake

AI-enhanced image of Minneapolis shooting shared online

A still image which has been taken from a real video of the shooting of Alex Pretti has been enhanced with artificial intelligence, resulting in an agent kneeling on the ground missing a head.

Evidence 4/5
Mystery 3/5
Jan 31, 2026 2
Rumor

Boots is not giving away mini perfume sets to those who complete surveys

Facebook posts and ads claiming the retailer is giving away a ‘premium mini perfume set’ to everyone who fills in a survey are not genuine.

Evidence 4/5
Mystery 1/5
Jan 31, 2026 1
Rumor

Posts on social media incorrectly name Green Party by-election candidate

Multiple posts on social media have falsely claimed that a former parliamentary candidate, Melissa Poulton, was chosen as the Green candidate in Gorton and Denton.

Evidence 4/5
Mystery 1/5
Jan 31, 2026 1
Rumor

JD Vance - JD Vance wrong that Minneapolis ranks No. 1 in unauthorized immigrants

"Right now we're focused on Minneapolis because that's where we have the highest concentration of people who have violated our immigration laws."

Evidence 2/5
Mystery 1/5
Jan 31, 2026 4
Rumor

AdversaRiskQA: An Adversarial Factuality Benchmark for High-Risk Domains

Hallucination in large language models (LLMs) remains an acute concern, contributing to the spread of misinformation and diminished public trust, particularly in high-risk domains. Among hallucination types, factuality is crucial, as it concerns a model's alignment with established world knowledge. Adversarial factuality, defined as the deliberate insertion of misinformation into prompts with varying levels of expressed confidence, tests a model's ability to detect and resist confidently fram...

Evidence 2/5
Mystery 3/5
Jan 31, 2026 2
Rumor

Jared Moskowitz - Under Trump, more FEMA disaster assistance applications are in limbo for longer

The Federal Emergency Management Agency’s “backlog of unanswered disaster assistance applications has exploded to the largest in its history.”

Evidence 2/5
Mystery 2/5
Jan 31, 2026 2
Rumor

On the Insecurity of Keystroke-Based AI Authorship Detection: Timing-Forgery Attacks Against Motor-Signal Verification

Recent proposals advocate using keystroke timing signals, specifically the coefficient of variation ($δ$) of inter-keystroke intervals, to distinguish human-composed text from AI-generated content. We demonstrate that this class of defenses is insecure against two practical attack classes: the copy-type attack, in which a human transcribes LLM-generated text producing authentic motor signals, and timing-forgery attacks, in which automated agents sample inter-keystroke intervals from empirical...

Evidence 2/5
Mystery 3/5
Jan 31, 2026 2
Propaganda

Donald Trump - Trump says the U.S. leads China ‘by a lot’ in AI. Experts say that lead is measured by a few months

In artificial intelligence, “We're leading China by a lot.”

Evidence 2/5
Mystery 1/5
Jan 31, 2026 2
Deepfake

Industrialized Deception: The Collateral Effects of LLM-Generated Misinformation on Digital Ecosystems

Generative AI and misinformation research has evolved since our 2024 survey. This paper presents an updated perspective, transitioning from literature review to practical countermeasures. We report on changes in the threat landscape, including improved AI-generated content through Large Language Models (LLMs) and multimodal systems. Central to this work are our practical contributions: JudgeGPT, a platform for evaluating human perception of AI-generated news, and RogueGPT, a controlled stimul...

Evidence 2/5
Mystery 3/5
Jan 31, 2026 2
Rumor

What’s driving down US overdose deaths? Experts cite drug supply changes, increased naloxone

What's driving down US overdose deaths?

Evidence 2/5
Mystery 2/5
Jan 31, 2026 2
Deepfake

A Marketplace for AI-Generated Adult Content and Deepfakes

Generative AI systems increasingly enable the production of highly realistic synthetic media. Civitai, a popular community-driven platform for AI-generated content, operates a monetized feature called Bounties, which allows users to commission the generation of content in exchange for payment. To examine how this mechanism is used and what content it incentivizes, we conduct a longitudinal analysis of all publicly available bounty requests collected over a 14-month period following the platfo...

Evidence 2/5
Mystery 3/5
Jan 31, 2026 2
Rumor

Democratic fight over DHS funding could lead to government shutdown. How could it play out?

Democratic fight over DHS funding could lead to shutdown

Evidence 2/5
Mystery 2/5
Jan 31, 2026 2
Deepfake

Explainable Deepfake Detection with RL Enhanced Self-Blended Images

Most prior deepfake detection methods lack explainable outputs. With the growing interest in multimodal large language models (MLLMs), researchers have started exploring their use in interpretable deepfake detection. However, a major obstacle in applying MLLMs to this task is the scarcity of high-quality datasets with detailed forgery attribution annotations, as textual annotation is both costly and challenging - particularly for high-fidelity forged images or videos. Moreover, multiple studi...

Evidence 2/5
Mystery 3/5
Jan 31, 2026 2
Rumor

A parent’s guide to measles: What to look for and how to prevent infection

Parents’ guide to measles: How to prevent infection

Evidence 2/5
Mystery 1/5
Jan 31, 2026 2