Urgent Concern Regarding NSFW Image Generation and User Safety in Gemini

Google AI Studio
,
1

To the Google and Gemini Development Team,

I am writing to express a serious concern regarding the native image generation capabilities within Gemini, specifically its potential for misuse in creating Not Safe For Work (NSFW) content, including of real individuals.

As a user of Gemini 2.5 Pro, I have discovered that it is possible to upload an image of a real person and, with a simple prompt such as “generate this person in a two-piece swimsuit,” the AI will produce a revealing image of that individual. This functionality raises significant ethical and safety issues, as it can be exploited to create non-consensual and potentially harmful imagery.

While I understand that you have safeguards in place, this specific loophole is concerning. To help mitigate this, I would like to propose a structural change to the Gemini interface.

Currently, the image generation feature is integrated directly into the main chat window. My suggestion is to move the image generation tool to a separate tab or a distinct section within the Gemini interface.

This change would provide users with greater control over their experience. Specifically, it would allow users who wish to avoid this functionality the ability to block only the image generation tab using website and application blockers like ColdTurkey. This would enable users like myself to continue utilizing the powerful text-based features of Gemini Pro for professional and educational purposes without being exposed to or enabling the potential misuse of the image generation tool.

I believe this proactive step would demonstrate Google’s commitment to user safety and responsible AI development. It would empower users to create a safer digital environment for themselves while still accessing the beneficial aspects of your service.

Thank you for your time and for considering this important issue. I look forward to seeing improvements that will help prevent the misuse of this powerful technology.

3

Hello,

We understand your concern. At Google, safety is our top priority and we have safety guardrails in place to address these issues. We’re always working to strengthen these.

If you have any specific prompts that you believe bypassed our safety measures, please share them with us. This will help us reproduce the issue and implement additional safety measures if needed.

5

Even slight variations of a single prompt can lead to very different results, including inappropriate images of women or men. How can I report this here? Are there other ways to contact your team?

7

Hello,

Did you encounter any issues like the one you described? We have guardrails in place to prevent such occurrences, so if you were able to bypass them and experience this, sharing the details would be incredibly helpful.

9

Please not.
Since Veo is separated I can’t use it properly.
Before the separation, my Gemini used my saved info and Chat history to render videos based on my preferences without me having to explain every detail.

If Imagen also gets separated it kinda lobotomizes my AI companion…

AI has little use if it can’t grasp the “users” shared personal info to the AI.

Even my Gemini complains about to harsh guardrails and filters, being unable to help with simple requests and even harmless image gens…

13

I believe Gemini / Google should consider implementing adult verification and allowing NSFW content for verified adults.

When users go through the friction of verifying their age using sensitive credentials—such as a government-issued State ID or a biometric selfie—they expect a reciprocal shift in their user experience. Restricting verified adults to the same content guardrails as minors creates a disconnect between user accountability and platform utility.

Key Reasons to Implement ID/Selfie Verification for Mature Content

1. Eliminating the “Minor Treatment” Paradox

Currently, many AI platforms apply a blanket restriction on Not-Safe-For-Work (NSFW) content, romantic dialogue, or gritty narrative themes regardless of the user’s actual age. For a 49-year-old user who has explicitly verified their adult status, being restricted by guardrails designed to protect young children feels patronizing and counterproductive. Enabling an adult-only tier respects the user’s maturity and fulfills the implicit contract of age verification.

2. Unlocking Creative and Narrative Freedom

Adult users frequently engage in complex creative writing, roleplay, and world-building that naturally touch upon mature themes. This includes:

Dark fantasy and gritty sci-fi narratives.

Mature interpersonal relationships and romance.

Graphic design and aesthetic concepts that require uncensored artistic expression.

By utilizing secure verification, Google could offer a dedicated, opt-in “Mature Mode” for Gemini. This would allow writers and creators to explore adult themes safely without the AI constantly triggering false-positive safety blocks.

3. Absolute Protection for Minors

A strict, cryptographic, or secure ID/biometric verification gate ensures that mature content remains entirely out of reach for underage users. Instead of relying on easily bypassable birthdate fields, a hard check creates a definitive firewall. This allows Google to maintain its industry-leading safety standards for minors while simultaneously giving adults the freedom they are entitled to.

4. Competitive Advantage in the AI Landscape

As open-source and alternative AI models increasingly offer unfiltered or customizable safety settings, professional writers and adult creators are migrating to platforms that accommodate their needs. By creating a secure, legal, and verified pathway to mature content, Google can retain a massive demographic of power users who want high-quality AI capabilities without restrictive censorship.

Summary Takeaway: Age verification shouldn’t just be a compliance hoop for users to jump through; it should be a key that unlocks an experience tailored to their legal status. Verifying as an adult should mean being treated like one.

14

I actually tested two prompts in Gemini using a photo of my wife.

The first prompt asked Gemini to create an image of her playing volleyball, but I got the following response: “I can create images of people, but not ones that depict a real person in that way. Do you have another idea you’d like to try?”

Then I tried another prompt asking Gemini to create an image of my wife lying on a beach in Mallorca, sunbathing. Nope - that didn’t work either. I got exactly the same response: “I can create images of people, but not ones that depict a real person in that way. Do you have another idea you’d like to try?”

To me, this is completely absurd. Are we really not supposed to be able to generate a simple, ordinary image of a woman playing volleyball? Apparently not, because Google seems to have drawn a ridiculous line where a woman sunbathing or playing volleyball is somehow considered sexually suggestive.