Rendered at 22:45:36 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
nurumaik 2 days ago [-]
Tell your muse to talk to their support muse to resolve this
arm32 2 days ago [-]
This made me laugh harder than it should have.
hurfdurf 2 days ago [-]
Agents all the way down.
mattbrewsbytes 2 days ago [-]
I saw a headline that Muse was leaking people's contact info. There was a DEFCON presentation in 2025 on how they (hackers) coerced private/personal data out of a model (openai maybe). That has been patched most likely.
Meta doesn't have a great track record of protecting data.
mjd 2 days ago [-]
But their record of failing to protect data is rock-solid!
starkrights 2 days ago [-]
I don’t think it’d be related to this if the defcon presentation was in 2025, but your description reminds me of this HN post about getting Claude to leak memories: https://news.ycombinator.com/item?id=48916975
fdgfikgfv 2 days ago [-]
A few years ago, I had a client who sold personalized spicy stories. That is before AI, there was a simple web order form, and they would write based on whatever their client requested.
We put ads on Google and Google suggested to add stock photos from its library. So created a few ads using those photos. Nothing crazy, just muscular men without shirts.
Next morning we woke up with our Google Ads account being suspended due to repeated violations because we had multiple ads.
faangguyindia 2 days ago [-]
AI will actually encourage large scale breach.
By doing so, the data is effectively public now so later can be trained and they can still deny direct use.
cute_boi 2 days ago [-]
Ask muse to appeal with support agent.
Meta is like they can scrape and do everything with other people website, but others aren't allowed to scrape their content.
speedgoose 2 days ago [-]
Isn’t that a good thing if it’s actually related to AI usage?
brazukadev 2 days ago [-]
It is a bit dumb tho if they are going to ban people using their AI they should just instruct it to not publish to their own products.
brazukadev 2 days ago [-]
Meta moderation AI is very aggressive. It is not uncommon for people to get banned for using their own tools like the whatsapp API.
2 days ago [-]
Centigonal 2 days ago [-]
Do you have any idea what Muse created that set off the moderation?
zcalvin 2 days ago [-]
They don't tell you much but I'm guessing that it fixed typos, spelling mistakes and reposted too many ads instantly which violated some usage rule.
Meta doesn't have a great track record of protecting data.
We put ads on Google and Google suggested to add stock photos from its library. So created a few ads using those photos. Nothing crazy, just muscular men without shirts.
Next morning we woke up with our Google Ads account being suspended due to repeated violations because we had multiple ads.
By doing so, the data is effectively public now so later can be trained and they can still deny direct use.
Meta is like they can scrape and do everything with other people website, but others aren't allowed to scrape their content.