AI Audit Disclaimer

This Disclaimer is a legally binding part of the AI Powered Smart Contract Audit Terms of Service. Read it before relying on any output of the Service.

1. This Is Not a Professional Security Audit

AI Powered Smart Contract Audit is an automated software tool that uses large language models and automated test execution to analyze smart contract code. It is:

Not performed, reviewed, or approved by a licensed, certified, or credentialed human security auditor.
Not a professional opinion, attestation, or certification of any kind.
Not legal advice, financial advice, investment advice, or a recommendation to deploy, invest in, transact with, or rely on any smart contract.
Not admissible or usable as a substitute for a professional audit opinion in legal, regulatory, insurance, or investment-diligence contexts.

If you are preparing to deploy a contract that will hold or control significant value, we strongly recommend obtaining a manual audit from one or more qualified, human security audit firms in addition to, and not instead of, use of this Service.

2. No Guarantee Regarding Absence of Vulnerabilities

A "CLEAN" verdict means only that our automated pipeline, within the scope and resource limits of a given run, did not identify and confirm a vulnerability matching its detection methodology. It does not mean the contract is secure, bug-free, or free from exploitable vulnerabilities.
The Service may produce false negatives (missing real vulnerabilities) and false positives (flagging non-issues). Findings are further filtered through automated compliance/skepticism/review stages, which reduce but do not eliminate either type of error.
Verification levels L0–L3 describe how much confidence the automated pipeline has already validated for a specific finding (e.g., whether tests compiled, passed, or demonstrated measured profit). They do not describe the completeness of the audit as a whole, and a report with no L3 findings does not mean no L3-class vulnerabilities exist in the code.

3. Scope Limitations

The Service does not analyze, and its verdict does not speak to:

Economic or oracle-manipulation risk beyond the specific tests generated and executed;
Governance attacks more sophisticated than first-order scenarios covered by the pipeline's architecture analysis;
Vulnerabilities located in imported third-party libraries or dependencies outside the audited scope, including transitively imported code excluded by the automatic scope filter;
Off-chain components, front-ends, key management practices, multisig configuration, upgrade/admin-key risk, or operational security;
Regulatory, tax, securities-law, or compliance status of the audited protocol or its token.

4. Assumption of Risk

By submitting code for audit or relying on any Deliverable, you acknowledge and agree that:

You are solely responsible for all decisions made using a Deliverable, including whether and when to deploy a contract, disclose a report to third parties, or represent a contract as "audited."
Any financial loss, exploit, hack, loss of funds, tokens, or digital assets, reputational harm, or other damage arising from reliance on a Deliverable — including a missed vulnerability or an incorrect "CLEAN" verdict — is at your sole risk, and is subject to the liability cap and exclusions in Section 9 of the Terms of Service.
You will not represent to any third party (investors, users, counterparties, regulators) that a Deliverable is equivalent to, or a substitute for, a manual audit by a licensed human security firm, without also disclosing this Disclaimer.

5. Public Disclosure of Reports

If you publish or share a Deliverable publicly or with counterparties, you must include this Disclaimer (or a clear, equally prominent reference to it) alongside the report, and must not selectively quote findings or the verdict in a manner that omits or obscures the AI-generated, non-professional nature of the audit.

6. Acknowledgment

Use of the Service constitutes your acknowledgment that you have read, understood, and accepted this Disclaimer in full.