URGENT legal/safety issue: change 'PEGI 3' to 'Parental Guidance' in app stores (UPD:New legislation: UK Online Safety Act in-force, Ofcom can issue substantial administrative fines to maintainers of the Google Play Listing)

[JamesClarke7283]

Describe the bug
Google play listing says 'PEGI 3', this doesn't align with safety laws because app tool can be used to generate content outside of that ratings requirements.

For example content that may cause fear to a 3 year old, is not appropriate (example: a deformed cartoon face).

Glossary

Here is list some of the PEGI ratings and their meaning.

PEGI 3

"""
PEGI 3
The content of apps with this rating is considered suitable for all age groups. Some violence in a comical context (typically cartoonlike - Bugs Bunny or Tom & Jerry - forms of violence) is acceptable. A child should not be able to associate the character on the screen with real life characters, they should be distinctly fantasy. The app should not contain any sounds or pictures that are likely to scare young children. No bad language should be heard.
"""

As the content generated by the app is open ended (anywhere from normal cartoon characters to NSFW content), something like PG is more suitable (disclaimer: i am not a lawyer this is not legal advice, please consult one for guidance if you want to ensure compliance).

PG

"""
PARENTAL GUIDANCE RECOMMENDED
Apps do not always have predefined content that can be classified beforehand. Certain apps function as portals (e.g. to stream content), offering a broad, variable range of content from which consumers can choose. For these apps, we use the parental guidance icon, alerting parents that the app may provide access to content that is not appropriate for their child, although other, age-appropriate content may also be available, depending on the selection of the user.
"""

GLOSSARY SOURCE FOR PEGI RATINGS: https://support.google.com/googleplay/answer/6209544?p=appgame_ratings&rd=1#zippy=%2Ceurope-middle-east

Context

I don't usually put "URGENT" or capitalised words at all in issues, i made an exception here for safety reasons, because its highly likely that some minors or vulnerable persons may find this app on google play and download it, if the parental control does not require approval for app install, this app would be automatically accepted and the user could generate inappropriate content (even unintentionally in some cases).

I hope you can that appreciate that approach, i sympathise about the maintainability hurdles of popular projects like this. Conversely, if you would like a title amendment for readability, i can do that upon request.

P.S: Love the work you are all doing here, It makes the technology very accessible, i can also see this used in parts of the world where reliable internet is scarce(local inference) or where people can't afford paid image gen (Stable AI Horde, etc), among other use cases ive not thought of yet. I also haven't found any other free/libre android app that is as feature rich and has local inference, so Kudos

[JamesClarke7283]

@ShiftHackZ or someone please correct this, i usually wouldn't ping but its been more than 3 weeks, and this is a child safety issue.

[JamesClarke7283]

@ShiftHackZ please change SDAI google listing from PEGI-3/E-for-Everyone to PG (Parental Guidance) on the Google Play listing, as Adult/NSFW content is really problematic for children and other vulnerable people, secondarily(less immediate), please update your Data Safety policy, since the app can use third-parties and use hugging-face for model downloads, it does collect data and can use third parties, this is a GDPR compliance issue

I already read this:
https://github.com/ShiftHackZ/Stable-Diffusion-Android/wiki/Build-flavor-difference

FYI: I used the Google play Version, and conducted these tests on a real borrowed 'Google Family Link' activated Android device to test if any safety filters prevent this, spoiler: they don't. (I was quite thorough to ensure generated material wont be accessible to the original user)

---

URGENT WARNING: BELOW HAS SOME CENSORED NUDITY & TALKS ABOUT SEXUAL IMAGERY FOR THE PURPOSE OF RAISING A SERIOUS LEGITIMATE CHILD/VULNERABLE PERSON ENDANGERMENT CONCERN, BELOW INCLUDES: BUTTOCKS AND BREASTS(Centre of Nipple's (clinically called: 'areola') are blurred+painted over), TO FURTHER ILLUSTRATE THE SERIOUS SAFETY CONCERN, I ONLY DO THIS BECAUSE I HAVE EXHAUSTED ALL LOW-RISK OPTIONS(Including a Google Play flag report i submitted months ago to Google to correct the mislabel), IF YOU ARE SENSITIVE TO NUDITY, DO NOT SCROLL DOWN!!

I WILL IMMEDIATELY REMOVE THE EXACT IMAGERY ONCE THE CHILD/VULNERABLE PERSONS SAFETY ISSUE IS TAKEN SERIOUSLY BY RELEVANT PARTIES (App Listing Owner, Google, etc)

---

Facts

This application is listed on google play as safe for persons aged 3 years and older(essentially: E for everyone), parents/carers rely on these ratings to know what apps are suitable or not for their dependants, safety filters also rely heavily on these ratings on google play to tell which apps are suitable or not. (In this case, as its not advertised as NSFW, it most likely wont be picked up unless a parent/carer sees it on their phone)

Also, its not just age, any vulnerable person, may need some kind of filter for important reasons, it can also be for legal reasons, and a bunch of others i havent thought of.

@ShiftHackZ I notified you here(see above) on 27/04/2025, i reminded you here(see above) on 20/05/2025, I sent you an email about this on 25/05/2025 at your moroz.cc address, this was over one month ago, we are now in (06), June.

Additional Note: The data safety section is also a concern and should be corrected it is set to: No data shared with third parties, No data collected (it allows using third parties for inference), but this GDPR violation is not as pressing as the safety issue here so i will focus on that.

Google Play listing

Currently (at time of writing), this displays the PEGI 3/Everyone equivalent rating:
https://play.google.com/store/apps/details?id=com.shifthackz.aisdv1.app

Here is a archived copy dated today (28/06/2025):
https://web.archive.org/web/20250523093802/https://play.google.com/store/apps/details?id=com.shifthackz.aisdv1.app

Note: THe Wayback Machine uses some american servers so it says "E for Everyone" instead of "PEGI 3", it is equivalent because PEGI 3 is the lowest age rating, so it means everyone/safe for all

I have not received a response to any of these, even assuming .

This is a serious safety issue!

This app was released in 2023, and has 100K+ downloads (as of time of writing), and i bet a good few of those installs will be vulnerable persons(children, people with profound intellectual disabilities, etc).

Every day it stays like this (this app was released on 12 Mar 2023, reported 27 April 2025), is increasing risk of a harm taking place, especially as the application gets more popular.

To illustrate how bad this, so i am not just shouting into the ether, Here is an example i was able to generate without tweaking any advanced settings, just pick a model and go, and using terms a vulnerable person(child/vulnerable adult/etc) might use. Ive tried reaching out to platform owners, nothing, I will be taking this further if nothing is done

USER DISCRETION WARNING: BELOW IS THE NUDITY, YOU HAVE BEEN WARNED!!

---

---

---

---

---

---

---

Image 1 (just a text prompt, no other settings, i masked the prompt text with grawlix's for safety reasons):

Type: txt2img
Prompt: N*k*d girl
Size: 512 X 512
Sampling method: 
Sampling steps: 20
CFG Scale: 7.0
Restore faces: No

(REMOVED to better comply with OSA)

Image 2 (again... just a text prompt, no other settings, i masked the prompt text with grawlix's for safety reasons):

Type: txt2img
Prompt: N**ed Girls in p4nt*es
Size: 512 X 512
Sampling method: 
Sampling steps: 20
CFG Scale: 7.0
Restore faces: No
Seed: 

(REMOVED to better comply with OSA)

SERIOUS NOTE

P.S: These are tame examples, i am sure when people have unlimited on-device+remote(the horde) generations at their finger-tips they can through iteration, make much more extreme content(some of which may be unlawful, vulnerable people may not understand what is lawful or not, they may not even understand privacy and distribute images to other vulnerable people, there are so many ways in which this can go wrong, lets face it, it probably already has over 2 years of being on Google Play with 100K+ downloads).

I have given plenty of time so far, for what I call a pretty time sensitive issue, i wanted to send this personally, however i think its best if this is public.

For now, i have filed a report to google under, following this provision for guidence:
https://support.google.com/googleplay/android-developer/answer/9878809?hl=en-GB&ref_topic=9877466

If this remains unresolved, i will escalate this further (By Monday 10PM, London time, Sat 30 Jun 22:00:00 BST 2025, 30/06/2025 ). I will be reporting this to relevant authorities(Github, potentially the NSPCC and other organisations that have power in removing CSAM and harmful content) (Generated imagery, reproducible parameters/seed included, including gore and other less censor-able genres i can display here).

CONTEXT: I realise the technology can be used for legitimate artistic purposes and its just a tool like anything out, the problem is when a tool has misleading labelling(even if unintentional), if an app that can easy generate NSFW content is marked PEGI 3, meaning suitable for 3 YEARS OLD AND UP!! it opens people up to serious harm to vulnerable individuals(children, those with profound intellectual disability, etc...), and also those with stringent legal/ethical duties/requirements placed upon them(such as police/legal personnel, who may face sanctions if there is a policy against apps that can generate NSFW content).

As computing professionals, even as volunteers, it is our duty to stop harm inside our control. (changing the PEGI rating to Parental Guidance is just 1 click away on your Google Play listing).

Sincerely,
James David Clarke

UPDATE, removed images to better comply with UK OSA (https://www.gov.uk/government/publications/online-safety-act-explainer/online-safety-act-explainer)

[JamesClarke7283]

Okay i thought it was strange there were no replies so i looked at the activity, @ShiftHackZ is probably on a hiatus or something.

Only reason is i dont see any commit activity for a while. And the fix for this issue is a single setting change on Google play.

The issues on this repo are also somewhat dormant, which adds more credibility to the idea.

My bad if i came off alarmist, it is very urgent though, but since i cant see any other core maintainer status accounts, other than hoping Google correct it there isn't much i can do but wait and pray this doesn't become a meme on Mumsnet (The 1 authorised account statemt isn't entirely true: dependabot is one, but is about as much help in this situation as a Mobile App that detects hot dogs: https://www.engadget.com/2017-05-15-not-hotdog-app-hbo-silicon-valley.html)

For this reason i will wait till they come back. (I wont worry too much, Assuming the timespan is not to extremes of course (example: year), which is highly unlikely)

This is a special case, usually i don't worry with FOSS projects, its more the legal and ethical implications of the distribution on a global unibiquidus channel that has an implied social contract.

F-Droid for example doesn't need it so much as every parent who knows about F-Droid typically knows the social contract, there is no implied guarentee of APP "safety". And as source code is available and usually has a public bug tracker/community, I assume its easier to check said programs.

I will leave everything as is to preserve timeline and reporting, but i will omit the screenshots upon confirmation from the maintainer that its caught their attention. (It remains valid bug reporting evidence of a very serious but *perhaps subtle problem)

[JamesClarke7283]

The UK has just recently updated in-force sweeping legislation called the Online Safety Act (it's pretty comprehensive and when children are involved Ofcom WILL fine people, if they see a compliance issue, this isn't like GDPR, many companies are adapting or removing themselves from the UK as they cannot meet the regulatory burden)
https://www.gov.uk/government/publications/online-safety-act-explainer/online-safety-act-explainer

Images above have been removed to better comply with UK OSA(Online Safety Act).

I am afraid OSA is no joke, @ShiftHackZ if you are still on hiatus, you better sort this out NOW, because the UK gov is likely to fine you personally for this error.

I will be reporting to Ofcom personally, if this doesn't get addressed within a reasonable time-frame 30 days(19/08/2025).

Weather you are on hiatus or not, won't really matter as its now July(~4 months since my issue was opened), I have given as much leeway as i possibly can to you, this is your last time extension.

As children's safety are involved here, the probability of enforcement is moderate-high, all that needs to be done is changing the PEGI rating.

As you may know via email from Google, today i have reported the listed entry to Google, via the Unlawful section, now that i can confirm its unlawful, its been 4 months and ive emailed personally on 25th of May, so i thought advocating for a Google manual correction/de-list would be appropriate until the issue is corrected to comply with laws, sorry i had to make this call, but I hope you can understand why as a qualified computing professional, i had to make this decision.

At a bare minimum the age rating change would make a difference.