Skip to content
Chainshift
Contact us

Agent Terms of Use

Effective Date: October 22, 2026

Article 1 (Purpose and Application)

(1) These Terms set out the conditions for using the AI chat feature (the "Agent") that Chainshift Co., Ltd. (the "Company") provides in the Chainshift Dashboard.

(2) These Terms form part of the Company's Terms of Use (the "Base Terms"). Matters not set out in these Terms are governed by the Base Terms, and where these Terms and the Base Terms differ with respect to the Agent, these Terms apply first.

(3) Terms not separately defined in these Terms have the meanings given in the Base Terms and the Privacy Policy.

Article 2 (Advance Notice Regarding Generative Artificial Intelligence)

(1) The Agent operates on the basis of generative artificial intelligence. The Agent's answers, analyses, summaries, and suggestions, and the analysis contained in files generated by the Agent, are generated by generative artificial intelligence.

(2) The Company indicates on the Agent screen and in the outputs generated by the Agent that they were generated by generative artificial intelligence.

Article 3 (Definitions)

The terms used in these Terms have the following meanings:

  1. "Agent" means the AI chat feature within the Dashboard that answers the user's questions and retrieves and analyzes campaign data to which the user's organization has access in order to provide results, together with its associated skills and download features.
  2. "Input" means the questions the user sends to the Agent, the conditions the user selects (campaign, period, AI model, etc.), and attached files.
  3. "AI Response" means the answers, analyses, summaries, and suggestions that the Agent generates and provides in response to Input, and the files containing them.
  4. "Source Data" means the organization's campaign data (visibility and citation metrics, the original text of collected AI answers, cited source URLs, etc.) that the Agent retrieves to generate AI Responses. Source Data constitutes output of the Services ("Output").
  5. "AI Model Provider" means a provider of an external artificial intelligence model that the Company uses to generate AI Responses, and a business that intermediates or hosts calls to such a model.
  6. "Connected Service" means an external service (Google Search Console, Google Analytics, etc.) that the user connects to the Agent.

Article 4 (Service Content and Scope of Use)

(1) The Agent provides the following features. Specific features and usage limits are subject to the subscription plan and the guidance on the screen.

  1. Questions and answers about AI search optimization (GEO) and how to use the Dashboard
  2. Retrieval and analysis of campaign data and explanation of the results
  3. File downloads of Source Data
  4. Suggestions for improvement based on the results of analysis

(2) The Agent retrieves only campaign data to which the user's organization has access. If the user leaves the organization or the user's access to a campaign is revoked, the user can no longer access conversations, AI Responses, or files that used that data.

(3) Features marked "Beta" or similar on the screen are provided on a trial basis and may not be released as official features. The Company may change Beta features without notice to the extent that the change does not affect users' rights, fees, or the processing of personal information.

Article 5 (Nature of AI Responses and Responsibility for Verification)

(1) Because AI Responses are generated probabilistically, they may contain inaccurate or missing content. AI Responses that present specific figures and sources may also contain errors.

(2) Source Data reflects results as of the time of collection, and AI Responses may not reflect changes after that time. Parts of an AI Response that explain causes or reasons may be inferences rather than facts confirmed by data.

(3) Before using an AI Response or sharing it with others, the user must check the query conditions and cited sources displayed with the AI Response and, where necessary, have it reviewed by a person.

(4) AI Responses are not professional advice in fields such as medicine, law, tax, or finance. The user must verify on its own whether its use complies with the laws applicable to its industry, such as laws on medical advertising and on labeling and advertising.

(5) The mention of a particular brand, competitor, or website in an AI Response does not mean that the Company recommends or endorses it or has a partnership with it. AI Responses are not the Company's official opinion.

(6) The improvements suggested by the Agent do not guarantee any change in exposure, citation, or ranking in AI search results.

Article 6 (Processing of Input and AI Model Providers)

(1) To generate AI Responses, the Company transmits Input, conversation history, and Source Data to AI Model Providers. The Company transmits them only through the businesses disclosed under the outsourcing of processing and overseas transfer sections of the Privacy Policy (each, an "Intermediary"), and the Intermediary forwards them to the AI Model Providers needed to generate the AI Response.

(2) The Company may change AI models and AI Model Providers, taking into account the quality, stability, and cost of the service. Before transmitting to a new Intermediary, the Company will amend and publish the Privacy Policy.

(3) AI Model Providers may, under the Intermediary's policies and their own policies, retain the data they receive for a certain period or use it to improve their models. The Company discloses in the Privacy Policy the Intermediary's retention and use policies and where they can be reviewed.

(4) If the user leaves a rating or feedback on an AI Response, the Company may use the relevant conversation for internal review to improve the quality of the Agent.

(5) Conversation records can be viewed only by the user who created the conversation. However, to the extent necessary for support at the user's request, incident response, error investigation, security checks such as prevention of misuse, and compliance with laws, the Company's authorized support staff may view conversation records.

Article 7 (Rights to Input and AI Responses)

(1) The user retains all rights to its Input. The user permits the Company to store, transmit, and process Input to the extent necessary for the Company to provide the Agent.

(2) The user must have the legitimate right to provide Input (including attached files) to the Agent and must ensure that Input does not infringe the rights of any third party.

(3) To the extent permitted by applicable law, the Company assigns to the user any rights the Company has in AI Responses, and the user may use AI Responses in its own business. However, Source Data, the original text of collected AI answers, and cited web documents contained in AI Responses are governed by Section 4 (Content Ownership and Responsibility) of the Base Terms. The user may process and incorporate Output into its own reports, dashboards, consulting deliverables, and the like and provide them to its end clients (including an agency acting for an advertiser), but may not, without the Company's prior written consent, redistribute Output in bulk as raw data without substantial processing, sell or publish it as a dataset, use it to develop a product competing with the Services, or use it to train an artificial intelligence model; where the user provides Output to its end clients, it must impose equivalent obligations on them and is responsible for their violations.

(4) Due to the nature of artificial intelligence, the same or similar AI Responses may be generated for other users in response to the same or similar Input. AI Responses generated for other users are not subject to the assignment under paragraph (3).

Article 8 (Prohibited Conduct)

(1) In addition to the conduct prohibited under Section 5 (Prohibited Conduct) of the Base Terms, the user may not use the Agent for the following purposes or in the following ways:

  1. Attempting to discover the Agent's instructions (prompts), settings, or internal tool configuration, or manipulating Input to circumvent safeguards and usage restrictions
  2. Attempting to access data of an organization or campaign to which the user's organization does not have access
  3. Extracting AI Responses or Source Data by methods not provided on the screen, such as automated queries or bulk downloads using a program
  4. Creating content or plans that manipulate the sources of answers given by search engines or artificial intelligence services through networks of sites or accounts disguised as mutually independent sources, false testimonials or reviews, coordinated false postings, or the like. Improving and distributing content based on the user's actual information does not fall under this item.
  5. Presenting an AI Response as if it had been written solely by a person, thereby misleading others
  6. Collecting or analyzing another person's personal information without consent, or attempting to identify a specific individual through Input such as images
  7. Using AI Responses to develop or train an artificial intelligence model that competes with the Agent or the services of AI Model Providers
  8. Violating the usage policies of AI Model Providers that the Company has notified in advance on the Agent screen or through these Terms

(2) If a violation of paragraph (1) is confirmed, the Company may restrict the user's use of the Agent. Except in urgent cases, before imposing a restriction the Company will notify the user or the organization manager of the reason and provide an opportunity to explain.

Article 9 (Input of Personal Information and Confidential Information)

(1) The user must not enter into the Agent personal information unrelated to the purpose of the analysis. Apart from the Source Data that the Agent retrieves as part of its functions, the user must take care not to directly enter personal information not needed to generate AI Responses (in particular, sensitive information and unique identification information) or separate confidential information (including trade secrets) of the user's company.

(2) Where the user enters another person's personal information as required for its business, the user must satisfy the legal requirements applicable to it, such as notice and consent. The Company will also perform the legal obligations applicable to the Company with respect to the processing of such personal information.

(3) The Company is not liable for damages arising from the user's failure to comply with paragraph (1) or (2), to the extent of damages having an adequate causal relationship with such failure. However, where the Company's intent or negligence contributed to the occurrence or expansion of the damages, the Company is liable in proportion to its degree of fault.

Article 10 (Storage and Deletion of Conversation Records)

(1) The Company stores Input, AI Responses, and execution conditions and results so that the user can view conversations again. Conversation records can be viewed only by the user who created the conversation, and the Company's support staff may view them to the extent necessary under Article 6(5).

(2) When 30 days have passed from the date the user archives a conversation, the conversation records, execution history, and generated files are automatically deleted. If the conversation is restored before 30 days have passed, the scheduled deletion is cancelled.

(3) A conversation for which the user requests permanent deletion becomes inaccessible immediately upon the request, and the conversation content, attachments, and execution records are destroyed within 30 days from the date of the request.

(4) Conversations that are not archived are retained until the user deletes them or the user's organization terminates the service.

(5) Files generated for download expire 30 days after the date of generation. The extent to which they can be generated again depends on the user's data access rights and whether the underlying data is still retained.

(6) Notwithstanding paragraphs (2) through (5), records that must be retained under applicable laws are stored separately, with the relevant items and periods specified in the Privacy Policy. Additional retention for investigating misuse or responding to disputes is carried out only to the extent that a lawful basis, the subject, and the period have been confirmed, and such records are not used for any other purpose.

(7) Conversations whose destruction has been completed under paragraph (3) cannot be restored. Deleting a conversation does not affect the campaign data collected by the Company.

Article 11 (Connection of External Services)

(1) The user may connect Connected Services to the Agent. Before connection, the Company displays on the screen the scope of data that the Agent will retrieve. Once connected, the Agent retrieves data within the scope permitted by the user, uses it in AI Responses, and transmits it to AI Model Providers in accordance with Article 6.

(2) The user must have legitimate authority over the Connected Service account used for the connection.

(3) Connected Services are subject to their own terms and privacy policies. The Company does not guarantee the accuracy or availability of data from Connected Services. The user may disconnect at any time, and after disconnection the Agent will not newly retrieve the relevant data.

(4) Features through which the Agent performs tasks in external services on the user's behalf obtain the user's confirmation before execution. The user is responsible for the results of tasks it has confirmed. However, this does not apply where the Company's system performed a task different from the scope confirmed by the user, or where damage arose from the Company's intent or negligence.

Article 12 (External Disclosure of AI Responses)

(1) Where the user discloses an AI Response, or material derived from it, outside its organization, the user must review the content before disclosure.

(2) Where the user discloses an AI Response or a file generated by the Agent outside its organization in its original form or in substantially the same form, the user must keep the generative artificial intelligence label inserted by the Company. Where it is difficult to keep the original label because the material has been edited or restructured, the user must place a label with the same meaning in a position readily noticeable to readers. In either case, the user must not conceal the fact that generative artificial intelligence was involved in its creation or cause it to be mistaken as having been written solely by a person.

(3) Where the user discloses Source Data, or metrics derived from it, outside its organization, the user must display "Source: Chainshift" or "Powered by Chainshift" in an identifiable location, and when quoting an AI answer, must also state the name of the AI service that generated it and the time of collection.

Article 13 (Fees and Credits)

(1) The Company currently does not deduct credits for use of the Agent.

(2) On the Starter plan, the user may send up to a cumulative total of 100 messages to the Agent. When the usage limit is reached, use of the Agent may be restricted. The usage limits of other plans are subject to the subscription plan and the guidance on the screen.

(3) If the Company intends to apply fees, such as credit deductions, to use of the Agent, the Company will amend these Terms in accordance with Article 17 and give advance notice of the features subject to deduction and the basis for deduction.

(4) The nature of credits and refunds are governed by Section 7 (Fees and Payment) of the Base Terms. In accordance with the Act on Consumer Protection in Electronic Commerce, Etc. of Korea, the user may withdraw (cancel) the purchase of unused credits within 7 days from the date of payment, in which case an amount corresponding to credits already used will be deducted.

Article 14 (Changes to and Suspension of the Service)

(1) The Company may change the Agent's features, AI models, and usage limits due to changes in AI Model Providers' policies, technological changes, costs, or other reasons. The Company will give notice of material changes unfavorable to users at least 30 days before they take effect.

(2) All or part of the Agent may be temporarily suspended due to a failure, service suspension, or access restriction of an AI Model Provider. In such case, the Company will promptly give notice on the screen or through announcements and will make efforts to restore the service.

(3) Beta features are not guaranteed to be officially released, and the Company does not warrant that Beta features are fit for a particular purpose or free of errors.

Article 15 (Limitation of Liability)

(1) The Company is not liable for damages arising from any of the following causes where there is no intent or negligence on the part of the Company:

  1. Damages arising from the user's use or sharing of an AI Response without the verification under Article 5(3)
  2. The results of marketing or business decisions made by the user on the basis of AI Responses
  3. Damages arising from causes that the Company cannot reasonably control, such as natural disasters or failures of AI Model Providers or Connected Services
  4. Damages arising from the user's failure to comply with Article 9

(2) To the extent permitted by applicable law, the Company's liability for ordinary damages arising in connection with the Agent due to the Company's ordinary negligence is limited to the greater of (i) the total service fees actually paid to the Company by the user's organization during the 12 months before the damage occurred and (ii) KRW 100,000. However, this limit does not apply to the Company's intent or gross negligence, breach of personal information protection or confidentiality obligations, or liability that cannot be limited under applicable laws.

Article 16 (Reports and Inquiries)

(1) If an AI Response is inaccurate or inappropriate, or if the user believes that an AI Response infringes another person's rights, the user may notify the Company at its inquiry email address (dev@chainshift.co). The Company will review the matter and notify the user of the result.

(2) A user who objects to a restriction on use under Article 8(2) may raise the objection through the channel in paragraph (1).

Article 17 (Consent to and Changes of These Terms)

(1) Consent to these Terms is given by the manager of the user's organization on behalf of the organization. The Company separately records, as the consent history, the manager who consented, the date and time of consent, and the version of the Terms consented to.

(2) Changes to these Terms are governed by Section 12 (Changes to the Terms) of the Base Terms. However, the Company will individually notify users by email of changes unfavorable to users at least 30 days before they take effect.

Addendum

These Terms take effect on October 22, 2026.

This English version is provided for convenience. In the event of any conflict between the Korean and English versions, the Korean version shall prevail.