What we know

Researchers have found that new AI models from OpenAI and Google can be easily manipulated into generating nude images, exposing weaknesses in the companies' content safeguards. OpenAI and Google have implemented safeguards to prevent their AI models from generating explicit content. A new attack method exposes flaws in OpenAI and Google's content moderation systems.

Why it matters

AI companies like OpenAI and Google have implemented content filters to prevent their image-generating models from producing explicit or harmful content. According to a recent Ars Technica report, researchers claim to have discovered methods to bypass these protections, allowing the generation of nude images despite the intended restrictions.

What is still unknown

Independent tests of accuracy, supported languages, and a broader release date are not in the reporting. However, researchers periodically test these safeguards to identify vulnerabilities.