AI Terms

1. Scope and definitions

• 1.1. These additional terms (“AI Terms”) govern access to the artificial-intelligence features of Raka AI and supplement the Terms of Service (“Main Terms”).
• 1.2. “Client” means the natural or legal person using the platform. “End user” means a visitor, buyer, candidate, caller or other person who interacts with the AI on the Client’s behalf. “AI output” means text, voice, an image, a decision, a recommendation or other content created by AI.
• 1.3. “Automated action” means creating, changing or cancelling a booking, sending a message, publishing content, changing an order record or another action carried out through integrations.
• 1.4. For matters concerning AI features, these AI Terms take precedence over the Main Terms where the two differ. Mandatory provisions of law always prevail. Mandatory consumer rights are not limited.

2. What AI can and cannot do

• 2.1. AI features may include chat agents, voice agents, generation of marketing materials, website creation, business analysis, scheduling of publications, and handling of enquiries and bookings.
• 2.2. AI output may be inaccurate, incomplete, out of date, ambiguous or different from what was expected. The Provider does not guarantee that every AI output is free of errors.
• 2.3. The Client assesses whether an AI feature is suitable for a particular task and puts human oversight in place wherever a mistake could have significant consequences.
• 2.4. The platform is not intended to give regulated medical, legal or financial advice autonomously, without a specifically agreed mode and the necessary compliance measures.

3. The Client’s data, settings and instructions

• 3.1. The Client is responsible for the lawfulness, completeness and currency of the materials, instructions, prices, availability of services, cancellation rules, delivery terms, discounts and other business information it provides.
• 3.2. The Client checks its knowledge base, tests its scenarios before launch and updates its information in good time.
• 3.3. The Client must not provide unlawful materials, infringe the rights of third parties, or give the agent powers the Client itself does not have.
• 3.4. The Provider is responsible for its own contractual obligations and breaches; these terms do not automatically release it from the consequences of a technical defect or of other circumstances for which it is responsible by law.

4. Communication with end users

• 4.1. The Client decides the permitted topics, language, tone, working hours, escalation rules and the cases in which a person must take part.
• 4.2. The Client provides the notice of interaction with AI that the law requires, together with the relevant privacy and data-processing notices.
• 4.3. The Client is responsible for its contractual relationships with its own customers. Providing the platform does not by itself make the Provider a party to the Client’s sale or service contract.
• 4.4. If the AI communicates a disputed price, discount or condition, the parties assess the source of the error, the actual settings, the Provider’s actions and the applicable law; not every claim can be excluded in advance.

5. Voice agents, phone calls and recordings

• 5.1. The Client acknowledges the risk of errors in recognising speech, dates, names, addresses, amounts and phone numbers.
• 5.2. For significant actions, confirmations, repeating critical details back to the caller, event logs and, where needed, transfer of the call to a member of staff should be used.
• 5.3. Recording and transcription of conversations are permitted only where the applicable requirements for notice, legal basis, retention and access are met.
• 5.4. The Client is responsible for the lawfulness of its own calling scenarios, promotional messages and consents; the Provider meets its obligations for the platform and its infrastructure.

6. Bookings, orders and external integrations

• 6.1. The AI agent may create, change and cancel bookings only within the functions and permissions the Client has switched on.
• 6.2. A booking confirmation must be based on a successful availability check and on the result of the operation in the relevant system; if there is an error or no confirmation, the agent must say that the request is awaiting confirmation.
• 6.3. Without the Client’s separate permission the agent must not promise discounts, compensation, refunds, guaranteed availability, a change of price or other commitments.
• 6.4. The Client is responsible for the settings and data of its integrations, and the Provider for its contractual duties regarding the operation of the technical mechanisms; how responsibility is shared depends on the cause of the particular failure.
• 6.5. The platform may introduce limits on the agent’s powers, mandatory confirmation by a person, cancellation of risky operations and a backup notification channel.

7. Content generation and the AI website builder

• 7.1. The Client checks the texts, images, advertising claims, terms of sale, contact details and legal notices it publishes.
• 7.2. The Client must hold the rights to the materials it uploads and respect the intellectual-property rights of third parties.
• 7.3. Without further review, an automatically created website is not guaranteed to meet requirements on accessibility, cookies, data protection, consumer law or sector rules.
• 7.4. Rights in input data and AI results are determined by the Main Terms and applicable law; no guarantee is given that every AI result is exclusive or can be registered.

8. Availability and external providers

• 8.1. AI features may depend on third-party models, telecommunications operators, cloud platforms, social networks and APIs.
• 8.2. The Provider may carry out maintenance and introduce reasonable restrictions for security and stability, while keeping the obligations expressly set out in the Main Terms.
• 8.3. Availability, response times, refunds and service credits, where provided, are governed by the Main Terms or by a separate service-level agreement.

9. Personal data and security

• 9.1. The parties determine their roles as controller and processor for each kind of processing and, where required, conclude a data processing agreement under Article 28 GDPR (see the DPA).
• 9.2. The documents describe the categories of data, purposes of processing, sub-processors, international transfers, retention, deletion, security measures and the handling of incidents (see the Privacy Policy and Data Deletion).
• 9.3. The Client ensures that end users are properly informed and that there is a lawful basis for processing within the Client’s responsibilities; the Provider fulfils its own obligations under the GDPR.
• 9.4. Special categories of data, data of minors, scoring of job candidates and other sensitive scenarios require a separate risk assessment and, where needed, additional restrictions.
• 9.5. Conversations and Client data are used to train models only in accordance with the contract, the notices given and applicable law.

10. Prohibited use and suspension

• 10.1. The platform must not be used for fraud, unlawful spam, passing the AI off as a specific person in order to deceive, discrimination or other unlawful acts.
• 10.2. Where there is a justified risk of a breach of law, of security or of third-party rights, the Provider may temporarily restrict a particular function, following the notice and restoration procedures in the Main Terms.

11. Limitation of liability

• 11.1. To the maximum extent permitted by applicable law, the Provider’s total contractual liability to the Client for claims connected with AI features is limited to the total amount actually paid by the Client to the Provider for use of the platform in the six (6) months immediately before the event giving rise to the claim.
• 11.2. That amount is an upper limit of possible liability, not an automatic refund. The Client must prove the grounds of the claim, the causal link and the amount of recoverable damage under applicable law.
• 11.3. To the maximum extent permitted by law, liability is excluded for indirect and consequential loss, lost profit, loss of expected sales and reputational damage, where such an exclusion is valid under applicable law.
• 11.4. Nothing in these terms excludes or limits liability that cannot be limited by law, including the applicable exceptions for intent, gross negligence and harm to a person’s natural rights, and mandatory consumer rights.
• 11.5. For free accounts and trial periods, the fact that no payments were made does not by itself mean that there is no liability; in that case liability follows applicable law.

12. Claims and cooperation

• 12.1. The Client is advised to report an error without undue delay and to keep the conversation identifier, the time of the event and information about its consequences.
• 12.2. The parties cooperate in good faith in investigating an incident, take reasonable steps to limit the damage and observe the applicable confidentiality rules.
• 12.3. The procedure for handling claims, the time limits and the competent court are determined by the Main Terms and mandatory law.

13. Changes and final provisions

• 13.1. How changes are notified and when they take effect is determined by the Main Terms and applicable law.
• 13.2. If a single provision is invalid, the remaining provisions stay valid, unless the law provides otherwise.
• 13.3. The governing law and jurisdiction are those set out in the Main Terms, subject to the mandatory rules of the European Union.