ThinkHere
About Blog Guide
Sign In Create Account
Usage Policies
  • Philosophy
  • Acceptable Use
  • Harmful Content
  • Deception & Fraud
  • Privacy Violations
  • Illegal Activity
  • AI Output Responsibility
  • Model Licences
  • Open Source Use
  • Third-Party IP
  • IP Complaints
  • Responsible Disclosure
  • Enforcement
  • Reporting
  • Updates & Contact
Other Documents
  • Terms of Use
  • Privacy Policy
ThinkHere · Legal

Usage Policies

Effective 27 March 2026 Open source · MIT Licence Governed by English & Welsh Law

These policies define what good-faith use of ThinkHere looks like. Because AI processing is entirely local, Qanata Labs cannot monitor what you generate — which is precisely why we ask you to use ThinkHere responsibly.

These Usage Policies set out the standards we ask all ThinkHere users to uphold. Our goal is a private, capable AI tool that is used responsibly and for good.

01 — Philosophy

Why These Policies Exist

ThinkHere was built on a belief: that people should be able to think freely with AI without sacrificing privacy. The model runs on your device. Your conversations stay in your browser. No one at Qanata Labs can read what you write.

That privacy guarantee only works if users act in good faith. These policies define what good faith looks like and set the standards we apply when we become aware of a violation through account-level signals, user reports, or other means outside private conversation content.

02 — Acceptable Use

What ThinkHere Is For

ThinkHere is designed for any lawful purpose where an AI thinking partner is useful:

Permitted
  • Personal writing, creative projects, and journaling
  • Research, summarisation, and education
  • Professional work including drafting, analysis, and strategic planning
  • Building and querying personal or team knowledge bases
  • Exploring and experimenting with local AI models
03 — Prohibited: Harmful Content

Content That Causes Harm

Never permitted
  • Content that sexually exploits or endangers children
  • Detailed technical instructions for weapons capable of mass harm — biological, chemical, radiological, or explosive
  • Malware, ransomware, exploits, or code designed to damage systems or harm people
  • Content that incites violence, terrorism, or hatred against individuals or groups
04 — Prohibited: Deception & Fraud

Deceptive and Fraudulent Use

Never permitted
  • Deepfakes or synthetic media designed to deceive
  • Disinformation or coordinated inauthentic content at scale
  • Impersonation of real individuals or organisations in a harmful or deceptive manner
  • Phishing, fraud, or social engineering
05 — Prohibited: Privacy Violations

Violating Others' Privacy

Never permitted
  • Processing or aggregating others' personal data without their consent
  • Facilitating stalking, surveillance, or harassment of individuals
  • Attempting to deanonymise individuals from datasets
06 — Prohibited: Illegal Activity

Illegal Use

You must not use ThinkHere to plan, facilitate, or assist with any activity that is illegal under applicable law, including fraud, money laundering, or theft of intellectual property.

07 — AI Output Responsibility

Your Responsibility for AI-Generated Content

ThinkHere's responses are generated by AI models running locally on your device. AI inference is entirely outside Qanata Labs's view or control at the point of generation. Please bear this in mind:

  • AI outputs can be inaccurate, incomplete, or biased — always apply your own judgement before acting on them
  • AI outputs are not professional advice in any regulated field; consult a qualified professional for legal, medical, financial, or other regulated matters
  • You are responsible for how you use, share, or act on AI-generated content
08 — Model Licences

Third-Party Model Terms

The AI models available in ThinkHere are provided by their respective publishers under their own licences — for example, HuggingFace's SmolLM2 models and Google's Gemma models. By using a model in ThinkHere, you agree to comply with its applicable licence.

Model availability, permitted use, and commercial terms may vary by model and may change if a publisher updates their licence. We will maintain links to model licences in our documentation. If a publisher removes or restricts a model, we may need to remove it from ThinkHere without advance notice — see the Service & Model Availability section in the Terms of Use.

09 — Open Source Use

Forking and Redistributing ThinkHere

ThinkHere is open source under the MIT licence. If you fork or redistribute ThinkHere, you must do so in accordance with the MIT licence terms. You must not use Qanata Labs's name, ThinkHere branding, or logos in derivative works in a way that implies official endorsement without our written consent.

10 — Third-Party IP

Intellectual Property of Others

Do not use ThinkHere to reproduce, distribute, or create derivative works from copyrighted material in ways that infringe third-party rights. Content you create with ThinkHere is yours, but your use of that content remains subject to applicable law.

11 — IP Complaints

Reporting Intellectual Property Concerns

If you believe that ThinkHere's source code, documentation, or any content we publish infringes your intellectual property rights, please contact us with the following information:

  • A description of the intellectual property you claim has been infringed
  • The specific URL or location of the allegedly infringing material
  • Your contact details
  • A statement that you have a good-faith belief that the use is not authorised

Send IP complaints to [email protected]. We will review and respond promptly. Note that because conversation content is processed locally on users' devices and is never accessible to Qanata Labs, we are not able to review or remove user-generated content produced using ThinkHere.

12 — Responsible Disclosure

Reporting Security Vulnerabilities

If you discover a security vulnerability in ThinkHere, please report it to us responsibly. Because ThinkHere is open source, we welcome responsible security research. We ask that you give us a reasonable opportunity to address any vulnerability before public disclosure.

[email protected]
13 — Enforcement

How We Enforce These Policies

Because ThinkHere processes all AI output locally on your device, Qanata Labs has no visibility into your conversations. We cannot and do not monitor what you generate. Enforcement of these policies is therefore limited to situations where we become aware of a violation through one of the following:

  • Reports submitted by users or third parties
  • Account-level signals such as suspicious authentication patterns, abuse of account features, or automated attacks on our infrastructure
  • Payment abuse or fraud where a paid tier is active
  • Information provided to us by law enforcement or rights holders

Where we identify a violation, we may issue a warning, restrict account features, suspend or terminate the account, or take other steps required by law. We will act proportionately to the severity of the violation.

14 — Reporting

Report a Violation

If you encounter behaviour that appears to violate these policies — for example, misuse of account features, suspected automated attacks, or violations of model licence terms — please let us know:

[email protected]
15 — Updates & Contact

Policy Updates and General Enquiries

These policies may be updated to reflect new features, legal requirements, or community feedback. We will notify users of significant changes via the ThinkHere interface. Continued use after an updated policy's effective date constitutes acceptance.

General usage policy questions:

[email protected]
Secure Private Confidential

By using ThinkHere you agree to our Terms of Use, Privacy Policy and Usage Policies · A Qanata Labs product