AI Supplemental Terms

Last Updated: September 28, 2026

Effective date: September 28, 2026. Version: 1.3.

These AI Supplemental Terms are part of the Terms of Service between you and First Gear AI LLC, a Texas limited liability company, and control for AI-related subject matter.

1. What the Product Is

The Service installs an operator agent, which you name during setup, that works on top of a dedicated Supabase project we provision inside our own Supabase organization, pay for, and run for you. The data in that project is yours. The agent drafts, organizes, retrieves, and prepares work under your direction. It is built on a hard rule: anything external, irreversible, or consequential is prepared as a draft and waits for a human to approve it. The agent runs on Claude through your own Anthropic subscription.

2. Accuracy of Outputs

AI OUTPUTS ARE PROVIDED “AS IS.” OUTPUTS MAY BE INACCURATE, INCOMPLETE, OUTDATED, OR CONTEXTUALLY WRONG, EVEN WHEN THEY APPEAR CONFIDENT AND POLISHED. YOU ARE SOLELY RESPONSIBLE FOR EVALUATING AND VERIFYING OUTPUTS BEFORE RELYING ON THEM OR ACTING ON THEM.

3. Hallucination

Large language models are probabilistic systems. They can state falsehoods confidently, invent details, misread numbers, and cite sources that do not say what is claimed. The presence of citations, figures, or specifics does not guarantee correctness. Treat outputs as a capable first draft from a fast assistant, not as verified fact.

4. No Professional Advice

Outputs are not legal, tax, accounting, medical, or financial advice, and using the Service does not create any professional-client relationship. Regulated-industry rider: outputs are not engineering judgments, architectural or engineering stamps, building-code compliance determinations, safety certifications, or inspection results. Where law requires a licensed professional’s review, judgment, or signature, obtain it. This applies with particular force to construction, where your obligations to permitting authorities, inspectors, and clients are yours alone.

5. Your Responsibilities as the Human in the Loop

You and your authorized users are the decision-makers. You agree to: review drafts before approving anything external (messages, filings, payments, publications); verify facts, figures, dates, and obligations that matter before relying on them; keep humans responsible for decisions about people; and not represent AI output as human-verified work when it is not.

6. Liability and Indemnity for AI Use

The limitation of liability in the Terms of Service applies fully to AI features and outputs, including the fees cap and the exclusion of consequential damages. First Gear AI has no liability or indemnification obligation for claims arising from your use of outputs, including reliance on unverified outputs, use of outputs in violation of these Terms, or decisions you make with them.

7. Third-Party AI Infrastructure

The agent runs on your own Anthropic subscription. The database it works on runs on Supabase, under our organization and our bill, which makes Supabase our subprocessor rather than a vendor of yours. Any provider’s availability, pricing, terms, usage policies, and model behavior are outside our control, and model updates on their side can change how the agent behaves. If a provider change breaks a platform function, we do not guarantee feature parity or a fix. Your use of Claude must comply with Anthropic’s terms and Usage Policy under your own agreement with Anthropic.

8. Credential Rules

Your Anthropic account, your connected tool credentials, and any API keys are yours. Never paste API keys, passwords, or recovery codes into a chat with any AI, including this one. Per-user sign-in only. Do not share seats. Rotate any credential immediately on suspected exposure. First Gear AI never asks for your credentials in chat. Treat any such request as fraudulent and report it to hello@firstgear.ai.

9. Automated Decisionmaking: What This Product Does and Does Not Do

Honest characterization: the Service is an operator that drafts and organizes under human approval. By design, its outputs do not take effect on the world, or on people, until a human approves them. It is not designed or offered to make, or substantially replace human decisionmaking in, “significant decisions” about individuals: decisions producing the provision or denial of financial or lending services, housing, education, employment or independent-contracting opportunities or compensation, or healthcare.

If you choose to use outputs in such decisions, you are the deployer and controller for that use, and you are responsible for the legal duties that attach. Those may include pre-use notices, opt-out and appeal mechanics, human-review rights, impact or risk assessments, and record-keeping under laws such as the California CCPA regulations on automated decisionmaking technology (obligations applying from January 1, 2027), the Texas Responsible Artificial Intelligence Governance Act, other state automated-decision and AI laws, and, if you operate there, the EU AI Act. Keep a human meaningfully in those decisions. The product is built to make that the path of least resistance.

10. Short-Form Acknowledgment

This is the acknowledgment you accept at onboarding and see in the dashboard:

“Your agent drafts; you decide. AI outputs can be wrong even when they sound right, and they are not professional advice. Nothing external happens without your approval. Never paste passwords, API keys, or recovery codes into any chat.”

11. Data Handling for AI Features

The dedicated project sits in our Supabase organization and we pay for it. The data is yours, and it is locked on your screen before it reaches that project, with a key only you hold. We can see that your account exists, when you signed in, and how much storage you use. We cannot see what you saved. We do not sell your data, advertise against it, or use it to train AI models. Your conversations with Claude sit with Anthropic under your own account and are governed by Anthropic’s terms, not ours. You can export everything at any time and take it with you, and if you leave we keep a locked copy for you at no charge, in storage only, so you can retrieve it later or rejoin. Only your key opens it. Tell us to delete it and we delete it within 30 days and clear backups within 90 days. See the Privacy Policy and the Data Processing Addendum for detail.

12. Changes

These Supplemental Terms are versioned and updated per the Terms’ notice process. Point-of-use notices in the product may be refined without notice where they only strengthen or clarify the disclosures above.

13. Governing Law

These Supplemental Terms are governed by the laws of the State of Texas, without regard to its conflict of law provisions. Venue for any dispute lies in the state or federal courts located in Travis County, Texas.

Contact Us

If you have questions about these AI Supplemental Terms, contact us at hello@firstgear.ai.

First Gear AI LLC
5900 Balcones Drive STE 100
Austin, TX 78731