2. Documents and data
Upload only information you may lawfully process and that is necessary for the purpose explained to the people concerned. Keep sources current and define who may access them. Do not place passwords or keys in instructions, documents or messages; credentials belong only in the mechanisms provided for that purpose.
Uses involving sensitive data, children or legally restricted information require an expressly agreed scope and the safeguards applicable to that use. They must not be enabled through a general configuration intended for public information.
3. AI interactions
Users must be able to recognise when they are interacting with an AI agent. Generated responses may be incorrect or omit relevant information; using documents as context does not eliminate that risk.
Do not present an agent as a real person, fabricate sources or use it as the sole decision-maker for high-impact rights or services, such as medical diagnosis, credit approval, employment or access to essential services. Such decisions require responsible people and the applicable professional and legal framework.
4. Actions and human oversight
Limit each connection’s permissions and maintain an accountable operator. Actions creating orders, bookings, payments or changes to information must have the required authorisation and rely on confirmation from the system that executes them. A generated statement is not a substitute for that confirmation.
Respect requests for human assistance. Do not configure instructions to hide errors, invent outcomes or bypass controls. Describe support hours and contact channels accurately.
5. Campaigns, incentives and communications
A campaign organiser must publish understandable rules and honour advertised dates, eligibility, rewards or benefits, limits and redemption terms. The organiser must obtain any regulatory approvals required by the campaign’s mechanics and territory. Where chance determines prize allocation, the applicable promotional gaming rules must be assessed before publication.
Do not use participation data for advertising without the required permission. Respect contact preferences, opt-out requests and connected-channel rules. Receiving a coupon or playing a game does not by itself authorise marketing communications.
6. Prohibited conduct
- Fraud, impersonation, deceptive manipulation, unlawful discrimination, harassment or exploitation.
- Unauthorised collection or disclosure of data, or use of documents, brands or content without sufficient rights.
- Unauthorised bulk messages, evasion of third-party restrictions, or automation that bypasses recipient preferences.
- Attacks, malware, secret extraction, attempts to cross organisational access boundaries or disable security controls.
- Manipulation of metrics, redemptions, identities, campaign results or evidence of completed actions.
7. Incidents and measures
Report suspected abuse or incidents to info@rondotap.com, providing information needed to investigate without attaching sensitive data or secrets. RondoTap may proportionately restrict an affected function where necessary to contain a risk or breach, in accordance with the contract and law. The customer will be informed as soon as appropriate unless legal or security considerations prevent this.
8. Relationship to the contract
These rules do not add features, integrations, certifications or service levels to a contracted plan. They do not exclude RondoTap’s own obligations or the customer’s mandatory rights. Third-party services remain subject to their applicable terms.