Need clarification in latest email about upcoming changes to "Abuse Monitoring"

Gemini API
1

Developers have just received an email about changes to “Abuse Monitoring” which will take effect in December 13 2024.

Are we to expect that the entire list of usage policy violations will start to be enforced on this day, and apps which generate unrestricted creative content (e.g. choose your own adventure games) will need to migrate entirely to a different service within 1 month? The official usage policy is quite strict and even explicitly forbids graphic violence and regular consensual sex, which are legitimate use cases in these games.

Alternatively, can there at least be an easy/free way to validate inputs before sending them to Gemini, so we can ensure we comply with the policy? Something like OpenAI’s Moderation Endpoint, for example.

Speaking of which, I had a horrendous experience with OpenAI’s customer service with evidence detailing that they wrongly banned me: Reddit - Dive into anything

I don’t want a repeat of that with Google Gemini.

1 Like
2

Okay, I remembered they have all those safety settings in the request. Presumably, in the worst case scenario if they consider something a violation, they’ll send me an email, and then I can modify my safety_settings to comply.

1 Like