AI Trust & Disclosure

Effective Date: May 1, 2026

Sqyro is an AI-powered paralegal built for licensed attorneys. The lawyer remains the lawyer. Sqyro drafts, researches, organizes, and prepares – but every output passes through your judgment before it reaches a client, a court, or the record. This page sets out the commitments we make to you, and the commitments we expect you to keep when you use the platform. It's written so you can forward it to a client, attach it to an engagement letter, or hand it to your state bar.

1. The lawyer is in charge

Sqyro is a tool, not a substitute. Every memo, draft, citation, and analysis Sqyro produces is a starting point that requires the supervising attorney's review before it leaves the firm. We design every workflow – drafting, research, intake, calendaring – to surface intermediate work for human review rather than auto-send. Attorneys remain solely responsible for legal advice, court filings, client communications, and the practice of law. Sqyro does not give legal advice, does not establish an attorney-client relationship with end users, and does not substitute for the judgment of a licensed lawyer.

2. Your client data is never used to train AI

Sqyro does not train AI models on your data, your clients' data, your matters, your documents, or your queries. Period. We do not sell client data, share it with third parties for marketing, or aggregate it into datasets sold to anyone. Data flows through AI systems for the purpose of producing your output, then it does not persist in those systems for training. This commitment is contractual and is reflected in our Data Processing Agreement.

3. Every legal claim cites a verified source

Sqyro is built on top of verified legal corpora – federal and state case law, statutes, regulations, USCIS forms, immigration administrative decisions, and dockets – not on a model's recollection of what the law might say. When Sqyro tells you what the law is, it shows you the underlying authority. When a citation cannot be verified, Sqyro flags it rather than presenting it as fact. Sqyro's research workflow includes explicit verification steps for cited cases, statutes, and regulations before they appear in a draft. Hallucinated citations are the single largest risk in legal AI; we treat the verification chain as a first-class product.

4. We design against hallucination

Sqyro's research and drafting workflows are governed by behavioral rules derived from real attorney post-mortems – including rules that require Sqyro to verify cited statutes before presenting them, to retrieve full case text before screening for relevance, to reformulate empty searches with fact-pattern keywords before declaring a question unanswered, and to distinguish factual lookups (filing fees, processing times) from legal analysis. These rules are continuously refined as new failure modes are identified.

5. We tell you what Sqyro did

When Sqyro completes a task, you see the trail: which sources it consulted, which tools it used, which citations it verified, which it could not. The agentic timeline in the chat surface makes the work auditable. You should never have to guess what Sqyro did to produce an output.

6. What Sqyro will not do

Give legal advice to clients, the public, or anyone other than the attorney using the platform
Train AI models on attorney-client privileged information
Sell, rent, or share your data with third parties for advertising or analytics
File documents, send communications, or take action on a matter without the attorney's affirmative review
Pretend to be a lawyer, a paralegal employed by a third party, or anything other than a software tool operated by your firm

7. Bar compliance and supervisory obligations

Most state bars now require attorneys to supervise the use of AI in legal practice (see, e.g., ABA Model Rule 5.3 on the supervision of nonlawyer assistance, and the growing body of state ethics opinions on generative AI). Using Sqyro responsibly means: reviewing every output before it leaves the firm, verifying citations independently when stakes warrant it, disclosing AI use to clients in accordance with your jurisdiction's rules, and not sharing matter information that the client has not authorized. Sqyro provides the audit trail you need to demonstrate supervision. The supervisory obligation is the attorney's, not Sqyro's.

8. Disclosing AI use to your clients

We encourage attorneys to disclose Sqyro's role to clients in their engagement letters. Sample language: “Our firm uses Sqyro, an AI-powered legal assistant, to support drafting, research, and case preparation. Every output is reviewed by an attorney before it is used in your matter. Sqyro does not train AI models on your information, and your data is not shared with third parties for advertising or marketing.” You are welcome to adapt this language to your firm's voice and your state's requirements.

9. Reporting concerns

If you believe Sqyro has produced an inaccurate citation, hallucinated a fact, or mishandled data, please report it to trust@sqyro.com. Reports go directly to the engineering and compliance teams and inform our behavioral rule updates. Security vulnerabilities should be reported to security@sqyro.com.

10. Updates

This page reflects our current commitments. Material changes will be communicated to active subscribers and posted here. The Last Updated date below indicates the most recent revision.