Thank you to the North Carolina real estate attorneys who joined our July 9 session, "Who Brought That Bot to the Meeting?" This recap distills the key points — the legal-tech landscape, the options on the market, the NC-specific ethics issues, and a framework you can use to evaluate any AI tool before it joins your next closing call.
Why do client conversations matter so much in a real estate practice?
Because every closing starts with a conversation. Intake, contract review, title and survey issues, and the dozens of judgment calls in between all happen in real time — and that's where a real estate attorney demonstrates expertise, earns client trust, and wins repeat and referral business from clients, agents, and lenders.
It's also where the practice is hardest. Real estate attorneys work under conditions that make careful listening genuinely difficult:
- High transaction volume and compressed due-diligence and closing timelines
- Many parties on a single deal — buyers, sellers, real estate agents, lenders, and opposing counsel
- Imperfect circumstances — calls squeezed between closings, details arriving late, and last-minute changes
Medical and sales professionals have had purpose-built tools to support their most important conversations for years. Legal professionals largely haven't — until recently. The question isn't whether AI belongs in these conversations. It's which AI is safe to bring into them.
What are the four levels of AI adoption for attorneys?
In the webinar we mapped a simple "journey to AI enlightenment" — four levels that describe how most attorneys actually adopt AI:
- The Seeker — "There must be a better way." Still identifying pain points and use cases.
- The Dabbler — "I'll use the AI in my toolbar." Relying on platform assistants like Microsoft Copilot or Google Gemini.
- The Operator — "I updated my tech stack with a few new tools." Combining general and purpose-built tools.
- The Architect — "I spend my free time vibecoding my own tools." Building custom solutions.
Wherever you fall, the same principle applies: the convenience of a tool matters far less than how it handles confidential client information. That's especially true at Level 2, where the "AI in your toolbar" was built for general business use — not for privileged legal conversations.
What are your options for AI in client conversations?
There are three broad categories of tool, and they are not interchangeable. Here's how they compare on the dimensions that matter to a real estate practice.
The takeaway: general-purpose tools can transcribe and summarize, but they weren't designed for the confidentiality and privilege obligations that govern a real estate closing.
What are the legal and ethical issues for NC real estate attorneys using AI?
This is where real estate practice gets specific. Several NC rules, opinions, and statutes bear directly on using an AI tool in a client conversation.
- NC Rules of Professional Conduct. Rule 1.1 (competence, including Comment [8] on the benefits and risks of technology), Rule 1.4 (communication and informed consent), Rule 1.6 (confidentiality), Rule 1.9 (duties to former clients), Rule 2.1 (independent judgment), Rules 5.1 and 5.3 (supervising lawyers and nonlawyer assistants — including paralegals and closing staff), and Rule 8.4(c) (misconduct).
- NC ethics opinions. NC 2024 FEO 1 addresses the use of artificial intelligence in a law practice — competence, confidentiality, supervision, and fees. NC 2011 FEO 6 addresses cloud computing and SaaS, requiring "reasonable care" to safeguard confidential information. At the national level, ABA Formal Opinion 512 (generative AI) and 477R (securing communications from inadvertent disclosure) are instructive.
- Consent and recording — the closing-specific trap. North Carolina is a one-party consent state under the Electronic Surveillance Act (N.C.G.S. § 15A-287); improper recording is a Class H felony. But real estate deals routinely pull in out-of-state buyers, sellers, and lenders — and all-party consent states (California, Florida, Illinois, Massachusetts, Pennsylvania, and others) may apply the moment one of those participants joins the call. A tool that records by default, without prompting for consent, can put you offside in exactly the multi-jurisdiction closings that are common in NC practice.
- Data security and biometrics. The NC Identity Theft Protection Act (N.C.G.S. § 75-60 et seq.) governs data security and breach notification. Illinois' Biometric Information Privacy Act (BIPA, 740 ILCS 14) can reach NC lawyers whenever an Illinois client or meeting participant is involved — and several general note-takers create voiceprints through speaker diarization.
- Privilege and case law. US v. Kovel (2d Cir. 1961) recognizes that an agent of the attorney can preserve privilege. US v. Heppner (S.D.N.Y. Feb. 2026) is the cautionary counterpart: AI tools whose terms permit training on your data can waive privilege and work-product protection.
The through-line: educate yourself and your clients about these issues before adopting any tool.
Why does AI architecture matter more than marketing?
Because architecture determines how much risk exists before a single privacy policy is applied. Vendor websites say "we protect your privacy." Vendor terms of service often tell a very different story — reserving the right to retain recordings, store transcripts indefinitely, or use your data to train models.
After Ambriz v. Google and US v. Heppner, the practical standard is clear: only tools built on ephemeral processing with contractually enforced zero-retention can reliably protect attorney-client privilege and work product. A tool that must create and store a full recording and verbatim transcript of a closing call is accumulating discoverable, subpoenable, breachable data by design. A tool that never creates that file in the first place carries far less risk before you've written a single internal policy.
Definition — Ephemeral processing: An architecture in which the tool analyzes a conversation in real time without creating or retaining a full audio recording or verbatim transcript, discarding partial data at the end of the conversation. Risk is minimized by design, not by after-the-fact policy.
How do I evaluate an AI tool before bringing it into a closing?
Use six lenses. In the webinar we walked through a due diligence framework that works for any AI tool an attorney is considering:
- Vendor expertise — Is it purpose-built for the legal industry? Does the vendor understand attorney obligations?
- Product architecture — How does it work? What data does it access, use, and store?
- Terms & conditions — What do the terms permit? Can the vendor train AI on your data? Do the terms match the marketing?
- Ethics implications — Is its use transparent to your client? Have you given notice and obtained consent? Do you retain independent judgment?
- Privacy & security standards — Does it hold certifications like SOC 2 Type II? Where is your data hosted? What else protects it?
- Industry reputation — Do bar associations or malpractice carriers support it? Has the vendor faced recent litigation?
What should NC real estate attorneys do next? Six key takeaways
- Understand your obligations under the NCRPC. The duty isn't merely to avoid harm — it's to understand how the technology works.
- Read the contract, not the marketing. Always review data retention, AI-training clauses, and deletion rights before adopting a tool.
- Architecture matters more than promises. After Ambriz and Heppner, ephemeral processing and enforced zero-retention are what reliably protect privilege.
- Consent disclosure is best practice. Disclose AI use to clients, update your engagement letters, and don't use record-by-default tools with participants in all-party consent states.
- A third path exists — legal-specific tools. Purpose-built Legal Conversational Intelligence is designed to satisfy the NCRPC, NC ethics opinions, and the Heppner framework — not merely marketed that way.
- We can't turn our backs on AI. Refusing to engage with AI can itself raise competence concerns under NCRPC 1.1 Comment [8]. The goal is to think critically, adopt responsibly, and get to "yes" with tools built for legal practice.
Ready to see it in a real closing conversation?
Querious is Legal Conversational Intelligence™ built for attorneys — designed to support your client conversations in real time while keeping client data private, privileged, and compliant with your North Carolina obligations.
Book a personalized demo → or try Querious free for two weeks →
This article is general information and does not constitute legal advice. Review the ethics guidance and consent laws applicable in your jurisdiction.
