Artificial Intelligence (AI) Policy

Effective date: 24 July 2026 · Last updated: 24 July 2026 · Version 1.0

This AI Policy explains how Rosnil Technology Private Limited ("Rosnil", "we", "us") uses artificial intelligence and machine learning ("AI") across our apps and services, what that means for your data, and the limits of what AI can do. It applies universally to every Rosnil app and is part of our Terms & Conditions and Privacy Policy.

1. What we mean by AI

"AI" here means machine-learning models and related techniques used to perform tasks such as recognising text in an image (OCR), detecting document edges, enhancing or cleaning a scan, classifying content, or generating suggestions. These are statistical systems: they produce likely results, not guaranteed-correct ones.

2. Where we use AI

Examples of AI features in our current apps include on-device text recognition (OCR), document edge detection and perspective correction, and image enhancement/clean-up. Each app's listing and in-app text describe the specific AI features it offers.

3. On-device first

Our default and strong preference is to run AI locally on your device. For on-device features, the image and text are processed on the device's own hardware (using platform frameworks such as Apple's and Google's on-device ML kits), and the underlying content is not transmitted to Rosnil or to a third-party AI service for that feature.

4. Your data & AI

For on-device AI features, your content stays on your device, consistent with our Privacy Policy. We do not receive the images or text processed by these features, and we do not build a profile of you from them.

5. No training on your content

We do not use Your Content to train, fine-tune, or improve foundation models — ours or anyone else's. On-device models ship pre-trained with the app; running them on your files does not send your files anywhere or teach a central model.

6. If cloud AI is ever used

If we introduce a feature that requires sending content to a cloud AI service (for example, a server-side capability that cannot run on-device), we will, before it ships:

• clearly disclose it in the app and in this policy;
• make it opt-in where the law or good practice requires consent;
• describe what is sent, to which provider, for what purpose, and for how long; and
• contractually require the provider not to use your content to train their models and to delete it after processing.

Until such a feature is announced, assume our AI runs on-device.

7. Our responsible-AI principles

Privacy by design

Process locally, minimise data, and never sell or repurpose your content.

Transparency

Tell you where AI is used and be honest about what it can and cannot do.

Accuracy & honesty

Present AI output as a draft, not an authoritative result.

Fairness

Avoid features designed to unfairly discriminate; work to reduce foreseeable harmful bias in the tasks we ship.

Human control

Keep you in charge — you review, edit, accept, or discard AI output.

Accountability

Provide a clear channel to raise concerns and we will act on them.

8. Limitations of AI output

AI features can make mistakes. OCR can misread characters, especially with poor lighting, handwriting, unusual fonts, or low-quality images; edge detection and enhancement can misjudge a page. AI output may be incomplete or incorrect and must not be treated as legal, financial, medical, or professional advice. Do not rely on AI output for high-stakes decisions without independent verification.

9. Your responsibilities

• Review and correct AI output before you use, submit, or share it.
• Ensure you have the right to process the content you run through AI features.
• Do not use AI features to process content you are prohibited from processing.

10. Prohibited AI uses

You must not use our AI features to create or facilitate unlawful, deceptive, or harmful material — including forged or fraudulent documents, impersonation, unlawful surveillance, or content that infringes others' rights. See our Acceptable Use Policy for the full list.

11. Human oversight & no solely-automated decisions

Our apps are tools that assist you; they do not make legal or similarly significant decisions about you on their own. You remain the decision-maker for what to do with any AI output.

12. Security

On-device AI keeps your content on your device, which is the strongest safeguard. Any data in transit for update checks or (if ever introduced) opt-in cloud features is protected with encryption in transit (HTTPS/TLS).

13. Third-party AI components

Our on-device AI features may use platform machine-learning frameworks provided by the operating system (for example, Apple and Google on-device ML). These run locally as part of the app. We will name any third-party AI service in this policy if and when a feature sends content off the device.

14. Changes

We will update this policy as our AI features evolve, revising the "Last updated" date and giving prominent notice of material changes — particularly before any change that would send content off your device.

15. Contact

Questions or concerns about our use of AI:
Rosnil Technology Private Limited
1st Floor, Shop No. 14, Palm Beach Residency, Akhandanand Saraswati Marg, Nerul West, Navi Mumbai, Thane, Maharashtra 400706, India
Email: rosnil@apeitnow.com