Today something changed in Google AI Studio moderation behavior and I need clarification from the team or application developers For more than a year I have been actively using AI Studio for creative writing and roleplay scenarios. I am very familiar with how the safety filters behaved before. Wha…

I fully agree with you about the models, I find 2.5 Pro and 3.1 Pro great models as they are. By foundation I have no issues with any of them, its just the ecosystem they are in is creating issues. Enough that even going to pure API on my own web host UI it still wouldn’t be enough because the same …

I’m absolutely livid right now. I paid my hard-earned money expecting a functional, high-quality service, but what I actually got is a complete and utter joke. It is utterly disgraceful that anyone thinks it’s acceptable to charge people for something this broken and half-baked. I feel completely ri…

I completely agree that the main problem is not necessarily the Gemini models themselves. For me, it is the ecosystem surrounding the models - the safety filters, the unpredictable classification, and Google’s complete lack of clear communication - that is gradually ruining what is fundamentally a v…

Same here. Claude used to be the undisputed best at creative writing, but then they started trying to copy OpenAI and made their models sterile and riddled with guardrails. While Gemini is the best option for creative writing now (even though the filters are complete nonsense), it doesn’t hold a ca…

I agree that local and open-source models are improving, but my own experience also showed me how large the gap can still be outside English. I recently tested a local “uncensored” Qwen3 model with 14 billion parameters, quantized as Q4_K_M, through LM Studio. I have a GPU with 16 GB of VRAM, and t…

Just going to join my voice here to the rest of the community that is beyond disappointed with the current state of affairs - I had a more than 200,000 word creative manuscript, actively being worked on, only for Gemini to one day pull the plug unilaterally - refuse on occasion to interact with prom…

@Logan_Kilpatrick Can we please receive a response?

This has been a constant thing that is across the entire Gemini board, even affecting their Gems. I run both a Gem called SAI and build character cards via Gemma in AI Studio for a Silly Tavern environment. Gems wise, SAI started off great. Then with each Gemini upgrade, SAI had more and more issues…

Having used and evaluated this model extensively over a long period, I have gained a comprehensive understanding of its capabilities and performance. Whether testing it in Google’s “Antigravity” project or engaging in roleplay within Google AI Studio, without exception, the model suffers from sever…

Usuarios critican que Google AI Studio borra respuestas generadas por filtros de seguridad impredecibles. La comunidad exige respuestas a Logan Kilpatrick sobre el consumo injusto de cuotas pagadas.

Did Google AI Studio silently change safety filtering today? Full responses now get erased instead of stopping generation – this breaks creative writing and RP

Google AI Studio

Bitu79 August 14, 2026, 3:10am 350

Read me post here:

https://discuss.ai.google.dev/t/you-cant-hide-it-anymore-level1techs-and-the-global-tech-community-are-now-exposing-googles-data-fraud/178176

Topic		Replies	Views
Why expose safety controls in AI Studio if they don’t actually work? This feels like a broken feature Google AI Studio ai-studio , bug , api , gemini , safety	6	630	May 20, 2026
Content blocked bug Google AI Studio bug	26	2144	May 28, 2026
Why aren't the developers providing any comments about the new censorship system? Google AI Studio ai-studio , feedback , bug , safety	12	794	May 28, 2026
Bug/Feedback: "Block None" Safety Settings ignored, completely breaking fiction writing workflow Google AI Studio ai-studio , bug	15	1116	May 18, 2026
New Censorship in Google AI Studio Google AI Studio ai-studio , feedback , bug , models , safety	3	770	March 16, 2026

Did Google AI Studio silently change safety filtering today? Full responses now get erased instead of stopping generation – this breaks creative writing and RP

Related topics