At MaruthLabs, we believe the safest AI is the one that respects your data sovereignty. Our safety philosophy is built on the architecture of privacy and control, and on being direct about what we do with your data.
Traditional AI safety often relies on server-side filters. Our approach is architectural: a model small enough to run where the data already lives does not need to send that data anywhere.
On-device deployment. When Madhuram runs locally on your phone, laptop, or your own server, prompts and outputs are processed entirely on your hardware. Nothing is transmitted to us and we have no visibility into it. This is the strongest privacy guarantee we can offer, and it is available as a custom deployment.
Hosted chat and API. When you use chat.maruthlabs.com or our hosted API, requests are processed on our servers and are recorded. We are explicit about this rather than implying otherwise: the details are in Principle 04 below.
Small language models (SLMs) like Madhuram are efficient, but like all LLMs, they can hallucinate facts. We are committed to transparency regarding model capabilities.
We recommend Madhuram for specific, well-scoped tasks rather than as a general-purpose model in production-grade pipelines. Users should always verify factual claims generated by the model against trusted sources.
We filter our training data to remove harmful, toxic, or dangerous content before the model ever sees it.
Furthermore, Madhuram is a strictly closed-source model. By retaining exclusive control over the model weights, we prevent malicious actors from stripping away safety guardrails or fine-tuning the model for harmful purposes, such as generating disinformation, malware, or non-consensual content.
We would rather state this plainly than bury it. When you use the hosted chat or the API, we record your requests and the model's responses.
Why. Usage metering, debugging, and investigating abuse. These are operational needs, and they apply to every account.
Training. We use API data to improve future models only if you opt in. The setting is off by default and can be changed at any time under Settings, API in the chat interface.
Deleted conversations. Removing a chat from the interface hides it from your account but does not erase it from our systems. If you need a conversation permanently deleted, write to us and we will do it.
On-device deployments are exempt from all of the above, because nothing reaches us in the first place.
If you discover a safety vulnerability or a way to bypass our safety training, please report it directly to our engineering team. We prioritize safety patches over feature releases.
[email protected]We're committed to responsible AI development. Reach out to discuss our safety practices.
Contact Safety Team