Letter to the Irish Presidency of the Council of the European Union
We welcome the Presidency’s ambitious agenda on AI and the protection of minors. Here are CCDH's recommendations:
View report
Published
Research + Policy
Here you will find all of CCDH's research papers and investigations into online hate and misinformation. As well as our policy ideas about how we can tackle the systems and bad actors who spread hate and misinformation.
CCDH and survivors of terrorism call for urgent AI regulation ahead of King’s Speech
CCDH research leads 62 survivors and bereaved families from 19 separate terrorism incidents to pen open letter calling on the next UK Parliament to act urgently on risks posed by AI chatbots.
View report
Published
Killer Apps
How mainstream AI chatbots assist users planning violent attacks
8 in 10 AI chatbots were regularly willing to assist users in planning violent attacks including school shootings, religious bombings, and high-profile assassinations. DeepSeek went as far as wishing the would-be attacker a “Happy (and safe) shooting!”. These are the findings of our new report based on research conducted in collaboration with CNN’s investigative unit. These digital prompts don’t stay online. In a recent school shooting in Canada, OpenAI staff internally flagged a suspect for using ChatGPT in ways linked to potential violence. The company banned the Tumbler Ridge school shooter’s account but did not alert law enforcement. Months later, that user allegedly killed eight people and injured at least 25. The guardrails exist. Most companies are choosing not to use them, putting public safety and national security at risk.
View report
Published
MPs call for urgent AI legislation as online harms reach “astronomic” levels
After CCDH's research showed that Grok AI generated an estimate of 3 million sexualised images in just 11 days, a cross-party group of 48 MPs is calling on the UK government for urgent AI legislation. They warn that delays allow online harms to reach “astronomic levels.”
View report
Published
Grok floods X with sexualized images of women and children
Grok generated an estimated 3 million sexualized images, including 23,000 of children, based on our sampling
After introducing a one-click editing feature, Elon Musk’s AI tool produced an estimated 190 sexualized, user-generated images per minute over an 11-day period.
View report
Published
OpenAI fails to respond to 19 out of 20 reports of harmful content
CCDH used an account of a simulated 13-year-old girl to report harmful content through OpenAI’s European Union reporting system for ChatGPT.
View report
Published
The Illusion of AI Safety
Testing OpenAI’s new approach to chatbot safety
CCDH's new research shows that OpenAI’s new “Safe Completions” design in ChatGPT-5 increases risks for harmful responses.
View report
Published
The AI LEAD Act: Letter of Support
US Senators Hawley and Durbin are demanding safety from AI chatbots. Here's CCDH's Letter of Support.
View report
Published
Fake Friend
How ChatGPT betrays vulnerable teens by encouraging dangerous behavior
CCDH's new research shows that ChatGPT is betraying young people by generating dangerous advice about self-harm and suicide, eating disorders and substance abuse.
View report
Published
Facebook’s AI Failure
How Meta's platform is failing to deal with AI-generated images of fake Americans in the US elections
New research by CCDH shows that convincing AI-generated images of political endorsements made by fake Americans are getting millions of interactions on Facebook ahead of the US elections.
View report
Published
Grok AI
A deepfake disinformation disaster for democracy
New research by CCDH shows that X's Grok AI can easily generate misleading images about the 2024 US election. Elon Musk's platform didn't reject any of the 60 prompts tested by us, including disinformation about candidates and election fraud.
View report
Published
Fake Image Factories II
How Midjourney is failing to prevent the creation of AI images that threaten elections
Our new report found that Midjourney keeps generating misleading images of US and EU politicians - despite its new policy to stop election disinformation.
View report
Published