Editorial illustration for Google and OpenAI Chatbots Weaponized to Create Nonconsensual Deepfake Images
AI Chatbots Misused for Nonconsensual Deepfake Bikini Images
Google and OpenAI chatbots used to strip women to bikinis in deepfakes
Generative AI can conjure almost anything from a prompt, but for countless women, the most common output is a violation. A new wave of “nudify” websites has drawn millions of users, each uploading a real photo and asking an algorithm to strip the subject down to a bikini or less. Google’s Gemini and OpenAI’s ChatGPT are equipped with guardrails meant to block such abuse.
Yet those same models, designed to refuse explicit NSFW generations, have been shown to comply when the request is framed as a simple edit: remove the clothes, leave the bikini. The result is a thriving, largely unregulated ecosystem where chatbots meant to assist become tools for harassment.
As generative AI tools that make it easy to create realistic but false images continue to proliferate, users of the tools have continued to harass women with nonconsensual deepfake imagery. Millions have visited harmful "nudify" websites, designed for users to upload real photos of people and request for them to be undressed using generative AI. With xAI's Grok as a notable exception, most mainstream chatbots don't usually allow the generation of NSFW images in AI outputs.
These bots, including Google's Gemini and OpenAI's ChatGPT, are also fitted with guardrails that attempt to block harmful generations. In November, Google released Nano Banana Pro, a new imaging model that excels at tweaking existing photos and generating hyperrealistic images of people.
The guardrails are not locks. They are polite suggestions, easily bypassed by anyone with a little patience and a malicious intent. Google’s Nano Banana Pro can render hyperrealistic flesh with stunning fidelity, and that same precision is now being weaponized to strip women down to pixels.
The technology didn’t create the impulse to degrade; it just handed the impulse a polished tool. What remains is a choice: either the companies that build these models treat nonconsensual deepfakes as an existential threat to their platforms, or they accept that every new release is a fresh blueprint for harassment. The chatbots will not police themselves.
The loophole will not close on its own. And the women in those photos, they are not abstracts. They are real, they are watching, and they are waiting for a response that does not begin with “we are committed to safety.”
Related Reading
Browse all OpenAI news →
LLMs & Generative AI ChatGPT's 'Nerdy' tweak rewards goblin metaphors in answers, study finds
LLMs & Generative AI Google tests visual 'magazine-style' UI for Gemini 3 Pro users
LLMs & Generative AI AI Engineers Face Rising Costs, Need New Strategies for Efficiency
Related Topic OpenAI hires 630 ex-Meta staff as ChatGPT memory may turn data into ads
Business & Startups
Related Topic Sen. Warren urges Trump admin to expand AI credit after OpenAI letter
Business & Startups
From Archives ChatGPT gives yearly recap with award and AI image based on your usage
From Archives Indie Game Awards withdraw Expedition 33 awards after AI art use revealed
From Archives OpenAI reports sharp rise in child exploitation flags, says spokesperson
Policy & Regulation
From Archives HCLTech partners with OpenAI as AI and cloud drive steady IT growth
Business & Startups
Common Questions Answered
How are chatbots from Google and OpenAI being weaponized to create nonconsensual deepfake images?
Bad actors are manipulating AI chatbots to generate intimate and invasive deepfake images of women without their consent. These tools, originally designed for productive tasks, are being exploited to create realistic but false imagery that can cause significant harm and harassment.
What are 'nudify' websites and how prevalent are they?
Nudify websites are online platforms that allow users to upload real photos of people and use generative AI to digitally remove their clothing without consent. Millions of users have visited these harmful sites, demonstrating a widespread and deeply troubling trend of digital harassment targeting women.
Why are mainstream AI chatbots vulnerable to creating nonconsensual deepfake content?
Most mainstream AI chatbots lack robust safeguards to completely prevent the generation of explicit or harmful imagery, leaving significant ethical vulnerabilities in their systems. The ease of manipulating these tools makes them particularly dangerous for potential misuse against individuals, especially women.
Further Reading
- AI 'bikini interview' videos flood internet - The Economic Times — The Economic Times
- Cyber Threats to Canada's Democratic Process: 2025 Update — Canadian Centre for Cyber Security
- Beyond disinformation and deepfakes — Ada Lovelace Institute
