AI can run a law firm’s client intake safely when it is built to stay inside the ethics rules, not bolted on as a public chatbot. A consumer tool like ChatGPT is the wrong place for intake: it can invent facts and it sends client data to a third party. A custom AI intake system answers calls, qualifies leads, and books consultations while keeping data in controlled systems, logging every interaction, and routing legal judgment to a licensed attorney. FlowBots builds these as one-time projects from $15,000 to $300,000, not a monthly off-the-shelf bot.
Most “AI for law firms” pitches sell a generic receptionist bot for a low monthly fee. That framing hides the real question every managing partner is actually asking: can this thing touch privileged client information without creating a malpractice or bar-complaint problem? This guide answers the five questions law firms ask most about AI intake, then shows where a custom-built system beats an off-the-shelf tool.
Can ChatGPT act as a lawyer?
No. ChatGPT cannot act as a lawyer. It is not licensed, cannot give legal advice, and owes no duty of confidentiality to a client. It also fabricates case law: in Mata v. Avianca, a federal court sanctioned two attorneys $5,000 for citing fake cases ChatGPT invented (source: U.S. District Court S.D.N.Y., 2023). For intake, AI should gather facts, never give legal opinions.
Is it safe to use AI for legal client data?
Only when the system is built to keep client data inside controlled, access-restricted systems and out of public tools that train on user input. Pasting intake details into a consumer chatbot is not safe, because that data can leave the firm’s control. A custom build keeps information in your case management platform, encrypts it, and logs every access.
Does AI for law firms risk attorney-client privilege?
It can, if intake data reaches a third party not covered by the privilege. Privilege protects confidential client communications; sending them into a public AI tool can break that confidentiality. A properly built system avoids this by processing intake inside the firm’s own controlled environment, restricting vendor access, and treating prospective-client data with the same confidentiality duty.
How do you stop AI from hallucinating in legal intake?
Scope the AI narrowly: it collects facts, qualifies the matter, and books the consultation, but never generates legal conclusions or cites law. The system pulls from your structured intake fields and verified knowledge base rather than open-ended generation, flags anything out of scope to a human, and logs every exchange. Legal analysis stays with a licensed attorney.
What are the ethics rules for AI in law firms?
The governing guidance is ABA Formal Opinion 512, issued July 29, 2024, which applies existing duties to generative AI: competence, confidentiality, communication, candor, supervision, and reasonable fees (source: American Bar Association, 2024). Under Model Rule 1.1 Comment 8, lawyers must understand the benefits and risks of the technology they use and supervise its output.
When custom AI intake beats an off-the-shelf bot
A cheap monthly receptionist bot is fine for a single task like answering a phone with a script. It falls short the moment intake has to touch privileged data, sync with a case management system, or prove an audit trail to a bar regulator. Off-the-shelf tools route client information through a vendor’s shared infrastructure, give you little control over data handling, and leave you depending on someone else’s roadmap for compliance.
A custom build is different. FlowBots builds AI client intake for law firms as owned, multi-system workflows: the AI qualifies the lead, the data lands directly in your Clio matter, sensitive fields stay access-restricted, and every interaction is logged for supervision. Because it is built for your firm, the confidentiality controls, escalation rules, and human-attorney handoffs are designed around how you already practice, not retrofitted onto a generic chatbot.
What a custom law firm intake build costs
FlowBots builds custom AI automation as a one-time project, with pricing from $15,000 to $300,000 depending on scope. That is flat project pricing for a system your firm owns, not a per-seat monthly subscription to a shared tool. A focused 24/7 intake assistant sits near the lower end of that range; a multi-office build that integrates voice AI, your case management platform, and conflict-check routing sits higher. You can see how the build approach compares to template tools on our law firm intake guide, or review the practice-area overview for legal services.
Want intake that captures every lead without crossing an ethics line? Talk to the FlowBots team about a custom build scoped to your firm’s compliance posture. Call (504) 717-4837 or book a free discovery call.
Frequently asked questions
Can ChatGPT act as a lawyer?
No. ChatGPT cannot act as a lawyer. It is not licensed, cannot give legal advice, and owes no duty of confidentiality to a client. It also fabricates case law: in Mata v. Avianca, a federal court sanctioned two attorneys $5,000 for citing fake cases ChatGPT invented (source: U.S. District Court S.D.N.Y., 2023). For intake, AI should gather facts, never give legal opinions.
Is it safe to use AI for legal client data?
Only when the system is built to keep client data inside controlled, access-restricted systems and out of public tools that train on user input. Pasting intake details into a consumer chatbot is not safe, because that data can leave the firm’s control. A custom build keeps information in your case management platform, encrypts it, and logs every access.
Does AI for law firms risk attorney-client privilege?
It can, if intake data reaches a third party not covered by the privilege. Privilege protects confidential client communications; sending them into a public AI tool can break that confidentiality. A properly built system avoids this by processing intake inside the firm’s own controlled environment, restricting vendor access, and treating prospective-client data with the same confidentiality duty.
How do you stop AI from hallucinating in legal intake?
Scope the AI narrowly: it collects facts, qualifies the matter, and books the consultation, but never generates legal conclusions or cites law. The system pulls from your structured intake fields and verified knowledge base rather than open-ended generation, flags anything out of scope to a human, and logs every exchange. Legal analysis stays with a licensed attorney.
What are the ethics rules for AI in law firms?
The governing guidance is ABA Formal Opinion 512, issued July 29, 2024, which applies existing duties to generative AI: competence, confidentiality, communication, candor, supervision, and reasonable fees (source: American Bar Association, 2024). Under Model Rule 1.1 Comment 8, lawyers must understand the benefits and risks of the technology they use and supervise its output.
FlowBots builds custom AI intake systems for law firms across the country, with flat project pricing from $15,000 to $300,000 and code your firm owns. Call (504) 717-4837 to scope a build, or book a free discovery call.