AI Training Data and Content Licensing Agreement

Your data is yours. We use it to do the work you ask for, not to build public models, unless you decide otherwise.

1. Scope and application

This Agreement defines the rights, intellectual property boundaries and usage permissions that govern user-generated content (UGC), telemetry and datasets processed by Sidiora Labs Inc. (“Sidiora Labs”) through sidioralabs.com and its products, including Paxeer X, Protention, Centra and Gideon, and Hyperfide (the “Services”).

It forms part of the terms you accept when you use the Services. If you have a signed enterprise agreement with us, any additional or stronger commitments in that agreement prevail.

2. Definitions

Inputs: prompts, instructions, files, code, repositories, messages, calls and other material you or your users submit to the Services.
Outputs: text, code, images, actions, reports and other deliverables the Services generate in response to Inputs.
Customer Data: Inputs and Outputs together, plus files stored in your workspaces.
Usage Data: technical telemetry about how the Services operate, such as latency, error rates and feature usage, that does not include the content of Customer Data.
Feedback: ratings, comments and suggestions you deliberately send us about the Services.
Training: using data to create, pre-train, fine-tune or otherwise improve the weights of a machine learning model.

3. Ownership of inputs

You retain all rights, title and interest in your Inputs. You grant Sidiora Labs a limited, non-exclusive, worldwide licence to host, copy, process, transmit and display Inputs only as needed to provide, secure and support the Services for you, and to comply with law. This licence ends when the Inputs are deleted from the Services, subject to backup expiry and legal holds.

4. Model training permissions and opt-out

Zero default public ingestion: your proprietary business data and private documents are excluded from public foundation model pre-training. We do not train on Customer Data by default.
Opt-in only: Customer Data is used for Training only if you, or an administrator for your organisation, explicitly opt in to a training programme. Each programme states what is used and for which models.
Opt-out guarantee: you may revoke training permission at any time through account privacy controls or by emailing compliance@sidioralabs.com. Revocation applies to all data from that point on, and we will exclude previously contributed data from future training runs.
Third-party model providers: where the Services call models operated by third parties, those providers are contractually prohibited from training on your Customer Data.

Models already trained on data contributed under an opt-in cannot always be fully “untrained”. Revocation stops all future use; it may not remove the statistical influence of past training runs.

5. Usage data, feedback and safety

Usage Data: we may use Usage Data, aggregated or de-identified so that it does not identify you or reveal Customer Data, to operate, secure and improve the Services.
Feedback: if you send Feedback, we may use it without restriction to improve the Services. Please do not include confidential material in Feedback.
Safety review: we may automatically screen Inputs and Outputs, and a limited number of authorised staff may review flagged items, to detect abuse, enforce our terms and meet legal obligations. Material retained for safety review is kept only as long as needed for that purpose.

6. Intellectual property of AI outputs

As between you and Sidiora Labs Inc., you retain full proprietary copyright and commercial ownership rights in all Outputs generated through paid subscriptions, and Sidiora Labs assigns to you any right it may have in those Outputs. Use of Outputs from free or evaluation tiers is governed by the terms of that product.

Because of how AI systems work, Outputs may not be unique, and other users may receive similar Outputs. Your ownership does not extend to Outputs generated for others, or to third-party material that appears in an Output. Copyright protection for AI-generated material varies by jurisdiction and may depend on the extent of human authorship.

7. Your responsibilities

You confirm you have the rights and permissions needed to submit your Inputs, including any personal data, to the Services.
You are responsible for reviewing Outputs before you rely on or publish them. Outputs can be inaccurate, incomplete or out of date.
You will not use the Services to infringe intellectual property, extract model weights or training data, circumvent safety measures, or build datasets by systematically harvesting Outputs in breach of the product terms.
You will not represent Outputs as human-generated where the law or a platform requires disclosure that content was produced by AI.

8. Public data, open source and research

Public ledgers: transactions and records written to the Paxeer X network are public by design and are not Customer Data under confidentiality commitments.
Open-source code: software we publish on GitHub is licensed under the licence in each repository.
Research: the Prometheus Papers and other research materials are licensed under the terms stated on each publication. Any data used in our published research is aggregated, de-identified, synthetic or contributed with explicit permission.

9. Indemnification safe harbor

Sidiora Labs Inc. will defend paid-subscription customers against third-party claims alleging that an Output, generated through the Services in accordance with our standard operating documentation, infringes that third party’s copyright, and will pay resulting damages finally awarded or agreed in settlement.

This indemnity does not apply where:

you knew or should have known the Output was likely to infringe, or prompted the Services to reproduce or imitate specific protected works;
the Input itself infringed, or you lacked rights to it;
you modified the Output, or combined it with material not supplied by us, and the claim would not have arisen otherwise;
you disabled, bypassed or ignored safety features, citation tools or usage restrictions; or
the claim concerns trademark use of an Output in commerce.

To receive this protection you must notify us promptly in writing, give us sole control of the defence and settlement, and provide reasonable cooperation. This section states our entire liability for third-party infringement claims relating to Outputs, subject to any limits in your subscription terms.

10. Concerns from rights holders

If you believe your work has been used in a way that infringes your rights, including in material we have published or in Outputs, follow the notice procedure in our DMCA and Copyright Policy. For other questions about data used in our models or research, write to compliance@sidioralabs.com.

11. Personal data

Personal data within Customer Data is handled under our Privacy Policy and, for the EEA, UK and Switzerland, our GDPR and Data Protection Notice. Business customers may request a data processing agreement.

12. Disclaimer

Except as expressly stated in this Agreement or a signed agreement, the Services and Outputs are provided “as is”. To the maximum extent permitted by law, Sidiora Labs disclaims all implied warranties, including merchantability, fitness for a particular purpose, accuracy and non-infringement.

13. Changes and governing law

We may update this Agreement. We will not reduce the training protections in section 4 for data already submitted without your consent, and we will give account holders at least 30 days’ notice of material changes.

This Agreement is governed by the laws of the United States and the state in which Sidiora Labs Inc. is incorporated, without regard to conflict-of-law rules, except where mandatory consumer protection law of your country of residence applies.

14. Contact

Questions about this Agreement, training permissions or the indemnity: compliance@sidioralabs.com.