Terms of Service

EntryDesk Terms of Service

These EntryDesk Terms of Service ("AI Agent Terms") govern the provision and use of the AI-powered automation and integration functionalities (the "AI Agent Services") offered by Dcard US Inc, a company incorporated under the laws of the State of Delaware ("DCARD"). These AI Agent Terms supplement and are incorporated by reference into the applicable Master Services Agreement ("MSA"), Subscription Order Form, or any other executed agreement between DCARD and Client.

1. Purpose and Scope

EntryDesk is a proprietary A2A (Application-to-Application) agent built on the Model Context Protocol ("MCP") framework. MCP is an emerging AI-agent interoperability standard designed to overcome the limitations of language models by enabling secure, structured, and permissioned interactions between AI assistants and external data systems, including local devices, enterprise databases, and third-party SaaS platforms.

Through a controlled and audited Gateway architecture, EntryDesk connects authorized enterprise SaaS systems with AI services in a secure, governed, and context-aware manner. This includes, but is not limited to:

• Querying and interacting with enterprise databases.
• Managing documents across cloud storage services.
• Extracting and transforming information from emails, calendars, and chat platforms (e.g., Slack).
• Performing system-level or web-based operations such as file handling, data parsing, or web debugging.
• Executing cross-application automations with full adherence to enterprise IT governance policies.

EntryDesk is designed to function as a unified A2A MCP Agent within the enterprise IT ecosystem, ensuring all AI interactions are subject to central access controls, audit logging, and policy enforcement. This enables enterprises to confidently integrate AI capabilities across SaaS applications without compromising data security, regulatory compliance, or IT governance.

2. Third-Party Services and Infrastructure

The AI Agent Services provided by EntryDesk may rely on various third-party technologies and platforms, including but not limited to:

(a) Foundational AI models and inference infrastructure (e.g., large language models operated by OpenAI, Google DeepMind, Meta, or Anthropic); and

(b) External SaaS applications and systems authorized and configured by Client (e.g., Google Workspace, Microsoft 365, Slack, Dropbox, Notion, etc.).

Client acknowledges and agrees that:

• AI Model Providers. The foundational models and their behavior (including generated outputs, response times, or embedded logic) are operated by third-party vendors. DCARD does not control the training, accuracy, or reasoning of these models and disclaims all liability arising from their performance or outputs.
• Integrated SaaS Services. EntryDesk serves as an orchestration and gateway layer under the Model Context Protocol (MCP) architecture, facilitating secure connections between AI agents and Customer-authorized third-party SaaS services. DCARD does not operate or manage those third-party services and is not responsible for their availability, performance, data accuracy, API reliability, access limitations, or security posture.
• No Warranties for External Dependencies. To the extent that any functionality of the AI Agent Services depends on the continued operation or correctness of third-party platforms or interfaces, including but not limited to SaaS application APIs, authentication services, or third-party hosting environments, DCARD makes no warranties and shall not be liable for failures, disruptions, or inconsistencies resulting from such external systems.
• Client Configuration Responsibility. Client is solely responsible for ensuring that appropriate permissions, API keys, tokens, scopes, and compliance measures are implemented when authorizing EntryDesk to access or integrate with its third-party environments.

For the avoidance of doubt, DCARD shall not be liable for any delay, failure, data inconsistency, or error resulting from: the acts or omissions of third-party SaaS providers; changes or deprecations in third-party APIs; unavailability of third-party services; or third-party security or privacy breaches beyond Provider's control.

3. Use Restrictions

Client and its Authorized Users shall not, and shall not permit any third party to, directly or indirectly:

• Use the AI Agent Services in violation of any applicable law, regulation, or third-party rights, including but not limited to data protection laws, export control laws, and contractual obligations governing the data or systems accessed via EntryDesk;
• Use EntryDesk to access, extract, transmit, or manipulate data from third-party SaaS systems beyond the access rights or scopes granted by the Client's IT administrators.
• Use EntryDesk to automate tasks or workflows that intentionally bypass, disable, or override enterprise IT security controls, audit mechanisms, or approval workflows.
• Introduce malicious code, engage in denial-of-service attacks, overload the system via abusive automation triggers, or otherwise degrade the performance or integrity of EntryDesk or any connected third-party services.
• Use AI-generated outputs from EntryDesk as the sole basis for making regulatory, financial, or operational decisions without appropriate human review, validation, or supervision.

4. Client Responsibilities

The Client acknowledges and agrees that it is solely responsible for the following:

• Ensuring that it has obtained all necessary rights, licenses, and consents to connect EntryDesk to any third-party systems, applications, APIs, or cloud services (collectively, "External Systems"). Client represents that such integrations do not violate any applicable terms of service or contractual obligations with third-party providers.
• Where EntryDesk allows for configurable or selectable AI model providers (e.g., OpenAI, Google, Anthropic), Client is responsible for: Selecting model providers appropriate to its business and compliance needs. Evaluating and approving any model behavior, privacy implications, and risk exposure. Reviewing and validating all outputs before using them in production or decision-making processes.
• Ensuring that any data input submitted to the AI Agent is accurate, lawful, and compliant with all applicable internal policies. Client shall independently validate and interpret outputs from the AI Agent before taking actions based thereon.
• Properly configuring all workflow triggers, automation rules, and downstream integrations within EntryDesk, and ensuring such configurations comply with Client's internal governance, data security, and regulatory requirements.
• Client shall be responsible for all fees incurred based on its use of EntryDesk, including: Charges based on the number of Authorized Users. AI model usage (e.g., token consumption, inference calls, compute time). Technical services related to integration with External Systems or custom configuration.

5. Data Handling and Privacy

• DCARD does not store, index, or retain any personal or sensitive data processed through the AI Agent Services unless explicitly stated in a data processing agreement (DPA).
• Client data submitted through the AI Agent may be transmitted to third-party AI model providers solely for the purpose of generating a response and is not used to train or improve any models.
• DCARD will use commercially reasonable security practices to protect data in transit and at rest.

6. AI Output Disclaimer

Client acknowledges and agrees that the AI Agent Services provided through EntryDesk are powered by LLMs that may produce content that is incomplete, imprecise, contextually inaccurate, outdated, or otherwise unsuitable for direct reliance. While EntryDesk employs mechanisms under the Model Context Protocol (MCP) - including contextual memory, prompt optimization, and workflow alignment - to enhance the relevance and accuracy of AI-generated outputs, no guarantees are made regarding output quality or appropriateness for any specific use case.

Accordingly, Client expressly agrees that:

• All AI-generated content, recommendations, summaries, or other outputs from EntryDesk are provided for informational purposes only and must not be used as the sole basis for any external communication, legal position, financial decision, or regulatory filing.
• Outputs must be reviewed, validated, and approved by qualified human personnel before being used to support decisions, automated processes, or any form of dissemination, whether internal or external.
• Client remains solely responsible for how it interprets, relies upon, or acts upon any AI-generated output. EntryDesk is a tool that assists, but does not replace, professional judgment or domain expertise.
• DCARD disclaims all express or implied warranties regarding AI-generated content, including but not limited to accuracy, non-infringement, merchantability, and fitness for a particular purpose. Use of AI outputs is at Client's sole risk.

7. Availability and Updates

The AI Agent Services are provided on an "AS-IS" and "AS-available" basis. DCARD may update, modify, or replace components of the AI Agent Services at any time without notice. No specific uptime or SLA guarantees apply unless explicitly stated in the AI Agent Terms.

8. Fees and Trial Period

The AI Agent Services are subject to separate billing terms, including:

• 14-day free trial period for new Enterprise Clients.
• Ongoing fees based on the number of Authorized Users, compute usage, and integration complexity.
• All billing and pricing terms are defined in the applicable Order Form or MSA.

9. Suspension and Termination

DCARD may suspend or terminate access to the AI Agent Services if:

• Client or Authorized Users breach these AI Agent Terms;
• Use of the AI Agent Services creates a regulatory, legal, or security risk;
• DCARD elects to discontinue the AI Agent Services.

Suspension does not relieve Client's payment obligations for the applicable term.

10. Limitation of Liability

To the maximum extent permitted by law, DCARD's total liability in connection with the AI Agent Services shall not exceed the total fees paid by Client for such services in the twelve (12) months preceding the event giving rise to the claim. In no event shall DCARD be liable for indirect, incidental, special, or consequential damages, or for actions or omissions of third-party AI providers.

11. Governing Law

The AI Agent Terms are governed by the laws of Taiwan (R.O.C.). Any disputes arising under or in connection with the AI Agent Terms shall be subject to the exclusive jurisdiction of the court of Taipei District Court.