AI Principles

How we use artificial intelligence — and what we don't do with your data.

Our approach to AI

MatterReady takes a conservative, attorney-first approach to AI. Every AI feature is designed to assist — not replace — the attorneys and staff who use the platform. AI outputs are always informational and never constitute legal advice.

We believe AI should reduce administrative burden without introducing risk. That means no autonomous decision-making, no hidden data sharing, and no opaque processing. Every AI feature requires human review before action is taken.

What we use AI for

MatterReady uses AI in a targeted set of features, each designed to save time on operational tasks:

  • Chatbot intake — An AI-powered chatbot on your website helps prospective clients describe their legal needs and collects structured intake information for your firm
  • Contact qualification scoring — AI analyzes intake submissions to help prioritize leads by practice area fit, completeness, and urgency indicators
  • Billing intelligence — AI surfaces billing anomalies, time entry patterns, and potential leakage from your Clio billing data
  • AI assistant — Staff and attorneys can ask operational questions about their firm's data (e.g., "How many new personal injury leads came in this week?") and get answers sourced from your MatterReady data

Your data is never used for training

MatterReady does not use Customer Content to train, fine-tune, or improve AI or machine learning models — ours or anyone else's.

This is a firm commitment, not a default setting. Your client intake data, matter details, billing information, and operational data are never used as training data. Period.

Our upstream AI provider (OpenAI) is contractually prohibited from using data sent through their API to train their models. We use the OpenAI API with data-use protections — not consumer products like ChatGPT — and our agreement explicitly prohibits model training on customer data.

AI providers

MatterReady currently uses the OpenAI API as its AI provider. We chose a provider that offers contractual no-training protections for API customers.

Key protections in our agreement with OpenAI:

  • Data sent via the API is not used to train or improve OpenAI models
  • Data is processed only to generate responses for the specific request
  • OpenAI acts as a data processor, not a data controller

If we add or change AI providers in the future, we will ensure equivalent contractual protections are in place and notify subscribers in advance via our subprocessor list.

Data retention with AI providers

Data sent to OpenAI via their API may be retained by OpenAI for up to 30 days solely for abuse and misuse detection, after which it is permanently deleted. This retention is required by OpenAI's API terms and applies to all API customers.

During this retention period, the data is not used for training, is not accessible to OpenAI employees except for abuse investigation, and is stored in encrypted form.

Permission scoping

AI features in MatterReady only access data that the requesting user is already authorized to view based on their role and permissions. AI does not create new access pathways or bypass existing access controls.

For example:

  • A staff member asking the AI assistant a question will only get answers based on data they can already see in the portal
  • The chatbot only accesses the firm's chatbot configuration and public-facing intake settings — not internal lead data or billing information
  • MCP (Model Context Protocol) API keys have explicit scopes that limit which data an external AI tool can access

Professional responsibility

We understand that attorneys have professional obligations when using technology that touches client data. MatterReady's AI features are designed to be consistent with:

  • ABA Model Rule 1.6 (Confidentiality) — We treat all client data as confidential. AI subprocessors are bound by contractual confidentiality obligations. Data is not used for training or shared with third parties beyond what is necessary to provide the Service.
  • ABA Model Rule 5.3 (Supervision of non-lawyer assistance) — AI outputs are always presented as recommendations requiring attorney review. Qualification scores, conflict signals, and assistant responses are never self-executing. Your attorneys maintain supervisory authority over every AI-assisted process.

We encourage firms to review their state bar's ethics opinions on AI and technology use. MatterReady is designed to support — not replace — the professional judgment required by these rules.

Contact

If you have questions about how MatterReady uses AI, our data handling practices, or need information for your firm's ethics review, reach out:

For our full legal terms, see our Terms of Service and Privacy Policy.