Clerks
Log inGet started

AI Use & Automation Policy

Last Updated: November 29th, 2025

This AI Use & Automation Policy (“Policy”) governs Customer's and Users' access to and use of all artificial intelligence features, automations, workflows, and related tools provided by Dropex, Inc., a Delaware corporation doing business as “Clerks” (“Dropex,” “Clerks,” “we,” “our,” or “us”) through the Clerks platform available at chatclerks.com (collectively, the “AI Features”).

This Policy is incorporated into and forms part of the Master Services Agreement (“MSA”). Capitalized terms not defined in this Policy have the meanings set forth in the MSA.

By accessing or using AI Features, Customer agrees to comply with this Policy.

1. Scope of AI Features

AI Features may include, without limitation:

  • Natural-language queries, chat-based assistance, and search
  • Automated drafting, classification, summarization, and insights
  • Workflow recommendations, task sequencing, and predictive suggestions
  • AI-assisted creation, modification, or organization of structured records (e.g., bills, invoices, transactions, customers, vendors, accounts)
  • Data extraction, normalization, and auto-fill functionality
  • Embedded AI agents or assistants
  • Any machine-learning or automation tools provided by Clerks

AI Features may evolve over time. Dropex may update, replace, or modify models, algorithms, prompts, tuning methods, or logic without notice, provided such changes do not materially degrade overall subscribed functionality.

2. Customer Responsibilities When Using AI Features

2.1 Verification Required

AI-generated outputs (“AI Outputs”) may contain errors, hallucinations, omissions, approximations, or inconsistencies.

Customer must review, validate, and verify all AI Outputs before relying on them, approving them, or submitting them to any accounting system, regulatory authority, counterparty, or third party.

2.2 Not a Substitute for Professional Advice

AI Outputs do not constitute accounting, tax, legal, financial, compliance, or other professional advice and are not a substitute for review by a qualified professional.

Customer must not rely on AI Outputs as the sole basis for any high-stakes decision.

2.3 Human Oversight Required

AI Features may assist with — but may not autonomously finalize — actions that:

  • create financial, tax, or legal obligations
  • modify business-critical accounting data
  • approve transactions, invoices, or journal entries
  • impact audits, compliance, or regulatory reporting

Customer must maintain meaningful human review and approval over all such actions.

2.4 User Behavior Impacts Output

AI Outputs may depend on:

  • the phrasing, clarity, and completeness of User prompts
  • the context provided
  • the quality, accuracy, and authorization of Customer Data

Customer is responsible for ensuring all prompts and inputs are lawful, accurate, appropriate, and properly authorized.

3. Prohibited Uses of AI Features

Customer must not use AI Features to:

3.1 Generate Harmful or Illegal Content

  • Fraudulent, deceptive, or misleading financial documents
  • Impersonation, manipulation, or misrepresentation
  • Deepfakes or deceptive AI-generated representations
  • Content that violates any applicable law or regulation

3.2 Endanger People, Property, or Business Operations

AI Features may not be used as the sole basis for decisions that:

  • approve or execute financial transactions
  • create binding accounting or tax determinations
  • generate regulatory or compliance filings
  • bypass internal controls, approvals, or audit trails

3.3 Abuse or Reverse Engineer AI Systems

Customer may not:

  • attempt to extract training data or model parameters
  • reverse engineer, deconstruct, or model-map Clerks' AI systems
  • use AI Outputs to train or benchmark competing models
  • circumvent AI safety controls, filters, or guardrails

3.4 Generate Unauthorized or Unlawful Automations

Customer may not configure or trigger automations that:

  • create false, misleading, or unauthorized accounting records
  • bypass required approvals or controls
  • misrepresent identities, authority, or intent
  • generate communications or filings without authorization

4. Data Handling, Privacy & Security

4.1 Processing of Customer Data

AI Features may process Customer Data solely to:

  • generate AI Outputs
  • provide, operate, maintain, and improve the Services
  • prevent misuse, fraud, or abuse
  • comply with applicable law

Dropex does not use Customer Data to train broad, non-Clerks foundation models unless Customer expressly opts in.

4.2 Sensitive Data

Unless expressly permitted and all required agreements are executed (e.g., a BAA or DPA), Customer must not submit:

  • Protected Health Information (PHI)
  • sensitive personal data of minors
  • unmasked payment card or bank account credentials
  • export-controlled or sanctioned data

4.3 Access Control & Security

Customer must:

  • protect credentials and restrict access to authorized Users
  • configure automations responsibly
  • maintain secure devices and environments

5. Accuracy, Reliability & Model Behavior

5.1 AI Outputs May Be Inaccurate or Variable

AI models may:

  • produce incomplete or incorrect results
  • omit context
  • vary across similar prompts
  • provide outdated or approximate information

5.2 No Guarantees

Dropex does not guarantee that AI Features:

  • are error-free or fully accurate
  • reflect real-time data in all cases
  • identify all inconsistencies or compliance issues
  • are appropriate for every use case

5.3 Customer Responsibility

Customer is solely responsible for decisions, actions, omissions, approvals, and submissions based on AI Outputs.

6. Automation Rules & Guardrails

6.1 Review of Automated Workflows

Automations must:

  • be configured carefully
  • include appropriate approvals and safeguards
  • not be used to bypass human review

6.2 Responsibility for Automated Actions

Customer is responsible for:

  • all automated actions within its Account
  • incorrectly configured workflows
  • unintended consequences of automation logic

6.3 Operational Integrity

Automations must not:

  • override safety or compliance constraints
  • manipulate accounting data unlawfully
  • create misleading records or filings

7. Restrictions on Competitive Use

Customer may not:

  • use AI Outputs to build or train competing products
  • benchmark Clerks' AI systems for competitive analysis
  • scrape or harvest system behavior
  • perform model extraction or inference attacks

8. Monitoring, Enforcement & Suspension

Dropex may:

  • monitor usage for compliance
  • restrict, filter, or block content
  • suspend AI Features or automations posing risk
  • rate-limit excessive activity
  • terminate access for violations
  • report illegal misuse to authorities

Repeated or severe misuse may result in termination under the MSA.

9. Customer Indemnification

Customer will indemnify and hold harmless Dropex from Claims arising from:

  • misuse of AI Features
  • reliance on incorrect AI Outputs
  • unauthorized or unsafe automations
  • violations of this Policy, the AUP, or applicable law

10. Changes to AI Features and Models

Dropex may update, replace, or modify AI Features, models, safety controls, or automation capabilities as part of ongoing improvement.

11. Updates to This Policy

Dropex may revise this Policy periodically. Material updates will be communicated in accordance with the MSA. Continued use of the Services constitutes acceptance of the updated Policy.