No. Tax positions depend on the full facts of a return, the specific code sections, the IRS's current administrative interpretation, and case law that moves continuously. A chatbot estimating whether a deduction is allowable or whether a penalty is abatable would be unreliable and the client would build expectations on it. The system prompt declines specific positions, declines outcome predictions, declines penalty abatement estimates, and routes everything substantive to the consultation.

Yes. Configure your content with the common notices (CP14, CP504, CP2000, CP504B, Letter 1058, Notice of Deficiency, Final Notice of Intent to Levy, CP90, LT11, etc.) and the bot reads the type from the prospect's description. Each notice has a different deadline and a different intake, and the bot routes accordingly. For ambiguous descriptions, the bot asks the prospect to read the notice number from the top right corner, which usually resolves the type.

These are the most time-sensitive cases in tax controversy and the bot treats them accordingly. Notice of Deficiency with a 90-day petition window routes to the same-week calendar with the controversy attorney, no exceptions. The system prompt is explicit that this deadline does not extend, that missing it forfeits the right to challenge in Tax Court (forcing the harder pay-then-sue route), and that the consultation needs to happen with time to prepare and file.

Yes. Each state has its own revenue department, its own procedural rules, and its own appeal timelines. Configure the bot with the states your firm covers and it routes accordingly. California FTB, New York DTF, Texas Comptroller, and Florida DOR each have different controversy frameworks. The bot also routes nexus and multi-state planning inquiries to the planning attorney rather than the controversy team.

Yes. Configure those as practice subareas in your content. Business tax controversy involves S-corp distributions, reasonable comp, accountable plans, and entity-level audits, each requiring different intake. International tax includes FBAR delinquency, Streamlined procedures, Section 965 transition tax, GILTI, and CFC reporting, with different forms and timelines. The bot routes by subspecialty rather than queueing everything to a generalist.

Carefully. OIC is heavily marketed and many prospects arrive believing they qualify when they do not, or convinced they do not when they actually do. The bot captures the basic financials (income, assets, monthly expenses) and the tax debt amount, explains the doubt-as-to-collectibility and doubt-as-to-liability frameworks at a general level, and books the consultation. It does not estimate likelihood of acceptance or quote a settlement number, because both are functions of the IRS's reasonable collection potential analysis.

The bot recognises criminal tax signals (CID contact, grand jury subpoena, eggshell audit posture, voluntary disclosure consideration) and routes to the senior criminal-tax attorney immediately. The system prompt is explicit that the bot does not discuss substance with prospects in active criminal investigation contexts; it captures contact information, confirms the urgency, and books the consultation. Substantive intake in criminal matters happens in the privileged setting, not in chatbot logs.

Pre-engagement conversations are not yet attorney-client privileged communications. For tax controversy specifically, this matters because IRS subpoena power is broad and chatbot logs would not be protected by Section 7525 federally authorized practitioner privilege either, since the engagement has not started. Most tax firms set short retention (30 days) on conversation logs and migrate substantive content into the matter file under privilege after engagement, which is the same posture as web-form intake.