Researchers say small changes to the prompt can still produce troubling content despite the fix
June 18, 2026
News Desk
bdnews24.com
The latest public version of ChatGPT can be manipulated into generating sexualised images and depicting scenes of graphic violence using a simple prompt, researchers have told the BBC.
The findings came from British AI security startup Mindgard, which discovered a way to make ChatGPT create graphic images by slightly altering a widely shared prompt originally intended to produce humorous results.
Images generated by AI showed women who had been killed or sexually abused
Researchers document jailbreak techniques that bypass ChatGPT safety filters to generate sexualized and violent images, raising…
El problema expone deficiencias en los sistemas de moderación de contenido de la inteligencia artificial de OpenAI
A newly discovered glitch in ChatGPT is sending shivers down users' spines. By using a simple prompt...
Campaigners ‘deeply concerned’ about response to prompts about suicide, self-harm and eating disorders
An investigation into how AI chatbots respond to violent requests produces "deeply disturbing" results, as OpenAI denies ChatGPT…
