Legal
1.1 This Policy sets the rules for using Sinora. It forms part of the Sinora Terms of Service (the "Terms") and uses the terms defined there. It applies to the Customer, to every Authorized User, to every Agent and Customer Application running in the Customer's Workspace, and to anyone who reaches the Service through them.
1.2 The Policy has two parts. Part A (sections 2 to 4) describes how Sinora governs its own use of AI. Part B (sections 5 to 12) sets out what the Customer must and must not do. The examples in this Policy are illustrations and not a complete list.
1.3 The Customer is responsible for making sure its Authorized Users know and follow this Policy.
2.1 You are using AI. The assistants, Agents and generation features in Sinora are powered by large language models and other machine learning systems, some operated by third parties. They are software. They do not understand, reason or exercise judgment as a person does, and they are not licensed professionals.
2.2 Outputs can be wrong. AI models generate output by predicting likely content. As a result an Output may:
2.3 Agents can act wrongly. An Agent that has been given permission to act may take an action the Customer did not intend, take the right action on the wrong record, repeat an action, or fail to act. Approval prompts and permission settings reduce these risks and do not remove them.
2.4 No professional advice. Outputs are not legal, medical, mental health, financial, investment, tax, accounting, engineering, safety or employment advice, and are not a substitute for a qualified professional.
3.1 No training on Customer Content. We do not use Customer Content or Outputs to train or fine-tune generalized AI models, and we do not authorize our managed model providers to do so.
3.2 Transparency about providers. We tell customers which providers power our managed models and where they process data, in our Privacy Policy and subprocessor list, and give notice of changes as the Data Processing Agreement requires.
3.3 Customer control. We give the Customer controls over which people, Agents and Customer Applications can use each Connection, over who can see what in the Workspace, and over the publication of Customer Applications, which requires approval of the specific version to be published.
3.4 Isolation. Each Customer's Workspace is separate. One customer's content is not used to answer another customer's requests.
3.5 Limited human access. Our personnel access Customer Content only for the purposes stated in the Terms.
3.6 Accountability. Sinora's Chief Executive Officer is accountable for this Policy. We review it at least once a year and when the law or the Service changes materially. Concerns about AI behaviour, safety or misuse may be reported to [email protected].
3.7 No guarantee of filtering. We and our model providers may apply safeguards that refuse or filter some requests and outputs. These safeguards are not complete, and their presence does not relieve the Customer of its obligations under this Policy.
Sinora provides a general-purpose tool. The Customer decides what to use it for, what data to give it, and what to do with what it produces. Where a law assigns duties to the person who deploys or uses an AI system for a particular purpose, the Customer is that person for its use of the Service.
5.1 Review before reliance. The Customer must have a person with suitable knowledge review each Output for accuracy, completeness and appropriateness before the Customer:
(a) relies on it for a business, financial, legal, compliance, safety or personnel decision;
(b) publishes it or sends it to anyone outside the Customer's organization;
(c) files it with a court, regulator or government body;
(d) deploys or runs code that it contains in a production system; or
(e) uses it in any decision about an identifiable individual.
5.2 Verify facts and figures. The Customer must check facts, numbers, quotations, citations and references in an Output against their original source. Figures in dashboards and reports built from Connections must be reconciled to the source system before they are relied on for financial reporting, tax, payment or regulatory purposes.
5.3 Supervise Agents. The Customer must:
(a) give each Agent only the permissions and Connections it needs;
(b) keep approval steps switched on for actions that are hard to reverse, including sending email or messages to outside parties, making or changing bookings, posting publicly, making payments or refunds, changing prices, inventory or advertising spend, and deleting or overwriting data;
(c) monitor scheduled and automated tasks and stop any that misbehave; and
(d) assign a named person responsible for each Agent that acts without step-by-step approval.
5.4 Internal rules. The Customer should keep its own written rules for AI use that are suited to its business and the laws that apply to it, and train its Authorized Users on them.
The Customer must not use the Service, or allow it to be used, to do any of the following.
6.1 Break the law or harm people
(a) Break any applicable law or regulation, or help anyone else to.
(b) Create, request or distribute child sexual abuse material or any content that sexualizes minors. We report this material to the authorities.
(c) Threaten, harass, bully, defame or incite violence or hatred against any person or group, including on the basis of a ground protected by human rights legislation.
(d) Promote or assist terrorism, violent extremism, human trafficking, self-harm, or the development or acquisition of chemical, biological, radiological or nuclear weapons.
(e) Create non-consensual intimate imagery, or sexually explicit content depicting a real person.
6.2 Deceive
(a) Commit or assist fraud, phishing, scams or impersonation.
(b) Generate disinformation, fake reviews or testimonials, fabricated endorsements, or content designed to mislead people about its source or about a matter of public importance, including elections.
(c) Create an image, audio or video that realistically depicts a real person saying or doing something they did not, without that person's consent and a clear label that it is AI-generated.
(d) Present an Output as written or reviewed by a human or a licensed professional when it was not.
6.3 Misuse personal information
(a) Collect, compile or process personal information without a lawful basis and any required consent.
(b) Track, monitor or build a profile of an individual without their knowledge, or locate or identify a person from an image.
(c) Use biometric identification or categorization, infer emotions of people at work or in education, or infer a person's race, religion, health, sexual orientation, political opinions or other sensitive characteristics.
(d) Score or rank people based on their social behaviour or personal traits in a way that leads to unfair or unrelated treatment.
(e) Submit the categories of restricted data listed in the Terms, including regulated health information and payment card data, without our written agreement.
6.4 Attack or abuse systems
(a) Create or distribute malware, or gain or attempt unauthorized access to any system, account or data.
(b) Use the Service to scan, probe or attack any network, or to run denial of service activity.
(c) Mine cryptocurrency, run proxy, relay or anonymization services, or use the Workspace as general-purpose hosting unrelated to the Service.
(d) Access or collect data from a website or service in breach of its terms, in a way that circumvents technical access controls, or at a volume that burdens it.
(e) Bypass, disable or attempt to defeat safety measures, filters, usage limits or metering in the Service or in an underlying model, including by prompt injection or jailbreaking.
(f) Extract model weights, or use Outputs to train a model that competes with the Service or with an underlying model, where that provider's terms prohibit it.
(g) Share an account, or sign-in access, among more people than the Plan allows.
6.5 Send unlawful messages
(a) Send a commercial electronic message without the consent, sender identification and unsubscribe mechanism required by Canada's anti-spam legislation, the United States CAN-SPAM Act and Telephone Consumer Protection Act, the European ePrivacy rules or any similar law.
(b) Send bulk unsolicited messages, or use purchased, rented or scraped contact lists.
(c) Use a Connection to a messaging, email, advertising or social platform in breach of that platform's rules.
6.6 Infringe rights
(a) Infringe or misappropriate anyone's copyright, trademark, patent, trade secret, privacy or publicity rights.
(b) Ask the Service to reproduce substantial parts of protected works, or to imitate a living creator's distinctive work in a way that misleads.
7.1 Not designed for high-risk uses. The Service is not designed, tested or certified for uses where an error could lead to death, injury, serious financial loss or the loss of an important right or opportunity. The Customer must not use the Service as the sole or determining basis for a decision in any of these areas:
(a) hiring, promotion, discipline, dismissal, task allocation or monitoring of workers;
(b) credit, lending, insurance underwriting or pricing for individuals;
(c) eligibility for housing, education, health care, public benefits or essential services;
(d) medical diagnosis, treatment or triage;
(e) law enforcement, immigration, border control or the administration of justice;
(f) the operation of critical infrastructure, vehicles, medical devices or any safety-critical system; and
(g) legal advice, or financial or investment advice, to members of the public.
7.2 Conditions for assisted use. The Customer may use the Service to assist a human decision-maker in the areas in section 7.1(a) to (c) only if: (a) a qualified person makes the final decision and can depart from the Output; (b) the Customer has assessed and tested the use for accuracy and for unjustified adverse effects on protected groups; (c) the individuals affected are told that AI is used, as the law requires, and are given a way to obtain human review; (d) the Customer keeps records sufficient to explain the decision; and (e) the Customer meets every obligation that applies to it as a deployer of AI under applicable law. Uses in section 7.1(d) to (g) are prohibited without our prior written agreement.
7.3 Automated decisions about individuals. Where the Customer uses the Service to make a decision based solely on automated processing of personal information, the Customer is responsible for informing the individual at or before the time of the decision, for giving them the information the law requires about the personal information used and the main factors behind the decision, and for giving them an opportunity to have the decision reviewed by a person. This includes the requirements of section 12.1 of Quebec's Act respecting the protection of personal information in the private sector and Article 22 of the GDPR, where they apply.
7.4 Recruitment. A Customer that uses the Service to screen, assess or select job applicants must make any disclosure in its job postings that employment standards legislation requires, including under the Ontario Employment Standards Act, 2000.
8.1 Tell people when they are dealing with AI. Where an Agent or a Customer Application interacts with individuals outside the Customer's organization, such as through a chat assistant, an email, a message or a booking exchange, the Customer must make it clear that they are interacting with an AI system, unless that is obvious from the context.
8.2 Label synthetic media. The Customer must label AI-generated or AI-altered images, audio and video that could be mistaken for authentic, and must not remove any watermark or provenance marking.
8.3 Disclose where required. The Customer must make any other disclosure about its use of AI that the law, a regulator, a professional body or a platform requires.
8.4 No attribution to Sinora. The Customer must not state or imply that Sinora wrote, reviewed, verified or endorsed an Output or a Customer Application.
9.1 The Customer is the operator and publisher of each Customer Application. Before making one available to anyone outside the Customer's organization, the Customer must: (a) review and test it; (b) confirm that the audience it is shared with is the audience intended; (c) make sure it does not expose credentials, Connection data or personal information to people who should not see it; (d) give its users any privacy notice and terms that the law requires; and (e) make sure it complies with this Policy.
9.2 A Customer Application must not collect sign-in credentials or payment card details, impersonate another organization, or distribute malware.
9.3 We may take down or restrict a Customer Application that breaches this Policy or creates a security or legal risk.
10.1 Where the Customer uses a model through its own provider account, the Customer must also comply with that provider's usage policies. Where those policies are stricter than this Policy, the stricter rule applies to that use.
10.2 The Customer's use of our managed models is also subject to the usage restrictions of the underlying providers that we pass through to customers, as published at https://sinora.ai/subprocessors.
10.3 The Customer must have the right to connect each Third-Party Service and must comply with its terms, including limits on automated access and on the use of its data with AI.
The Customer must: (a) protect the email accounts used to sign in; (b) remove access promptly when a person leaves or changes role; (c) not paste credentials or secrets into chats, and use the Connection features to store them; (d) treat content an Agent reads from outside sources as untrusted; and (e) report any suspected security incident, vulnerability or misuse to [email protected] without delay. Good faith security research must be coordinated with us in advance.
12.1 Monitoring. We do not routinely review Customer Content. We may use automated signals and may investigate where we have reason to suspect a breach of this Policy, a security risk or unlawful activity, and the Customer will cooperate with a reasonable investigation.
12.2 Enforcement. If we reasonably believe this Policy has been breached, we may, as we consider appropriate to the seriousness of the breach: (a) warn the Customer; (b) remove or disable content, a Customer Application, an Agent or a Connection; (c) limit or suspend an Authorized User or the Workspace; (d) terminate the Agreement under the Terms; and (e) report the matter to law enforcement or a regulator and preserve and disclose relevant information as the law permits. We will give notice and a chance to remedy where that is practicable, and may act immediately where there is a risk of harm, a security risk or a legal requirement.
12.3 No refund. No refund is owed where we suspend or terminate for a breach of this Policy.
12.4 No duty to enforce. We are not obliged to monitor or enforce this Policy, and a failure to act in one case is not a waiver. We are not liable to the Customer or anyone else for the Customer's breach of this Policy.
12.5 Reporting misuse. Anyone may report suspected misuse of Sinora, or content from a Sinora workspace or application that they believe is unlawful or infringes their rights, to [email protected]. A copyright complaint should identify the work, the material complained of and where it is, the complainant's contact details, and a statement of good faith belief that the use is not authorized. We will forward notices of claimed infringement as required by the Copyright Act (Canada).
12.6 Changes. We may update this Policy as the law, the technology and the risks change, on the notice stated in the Terms.
Questions about this Policy: [email protected]. Reports of misuse or AI safety concerns: [email protected]. Security: [email protected].