AI Marketing for Criminal Defense Attorneys

There is a question people ask AI tools that they will not ask anyone else, and it precedes some of the most valuable cases in criminal defense. A detective left a card. A target letter arrived. An investigator called about something at work. The person receiving that contact is not ready to call a lawyer, because calling a lawyer feels like an admission. So they open ChatGPT and ask whether they have to talk, what a target letter means, and whether they need an attorney. The tool answers, generally accurately, and almost always advises consulting counsel before speaking to investigators. That is the single most valuable piece of advice anyone will give that person, it is being given for free, and the firm present in that conversation has an enormous advantage. This guide covers how to be there.

1The Pre-Charge Query Opportunity

Pre-charge questions are close to a perfect fit for AI tools, and that makes them the most important AI opportunity in criminal defense.

The person asking has not been arrested. They have no case number, no court date, and nothing that forces them to act. What they have is a frightening ambiguity and a strong reluctance to escalate by calling a lawyer. Asking a machine costs nothing, discloses nothing, and does not feel like admitting there is a problem.

These are also, consistently, the highest-value matters in a criminal practice. Investigations that reach the target-letter stage frequently involve white collar exposure, professional consequences, and clients with the resources to pay for serious representation.

• Privacy is the whole reason. No disclosure, no admission, no record. That is exactly what someone under investigation wants.
• The tools advise getting counsel. For anything involving law enforcement contact, the responsible answer is consult an attorney. Your prospect is being qualified for free.
• The questions are researchable. Procedure, rights, and what specific contacts signal are documented, which these tools handle well.
• There is no urgency forcing a search. Unlike an arrest, an investigation produces no court date. AI research may be the entire pre-decision phase.
• Competition is minimal. Most criminal firms produce charge content and nothing about the investigation stage.
• Family members ask too. A spouse who learns their partner is under investigation asks the same questions.

2AI as a Charge Triage Tool

The second major use case is severity assessment. Someone charged with an offense they do not understand asks whether it is a felony, what the penalty range is, and whether they need a private attorney.

For a portfolio practice this matters because it happens across your entire charge library. Every offense you handle generates these questions, and your charge content is either feeding those answers or it is not.

• Classification questions dominate. Is this a felony, what class, what does that mean for me.
• Penalty range questions follow. Maximums, minimums, probation eligibility, and whether jail is likely.
• Collateral consequence questions are heavy. Employment, licensing, immigration status, housing, and firearm rights. Often the client's real concern.
• State specificity is required. Criminal law is state law. Tools that get your state's classification wrong send people bad information.
• The public defender question appears here. People ask whether appointed counsel is adequate, and the answers vary in quality.
• Your charge library is the raw material. Accurate, state-specific, well-structured charge pages are what these tools can extract from.

3When AI Gets It Dangerously Wrong

This section matters more in criminal defense than in any other practice area we work in, and it is a reason to publish rather than a reason to ignore the channel.

These tools sometimes give confidently wrong information about criminal procedure. Penalty ranges from the wrong state. Outdated classifications after a statute was amended. And most seriously, reassuring answers about talking to investigators that would harm someone acting on them.

A person who reads that they should cooperate and explain themselves, and does so without counsel, can damage their position permanently. Firms publishing accurate, clearly attributed content on these questions are correcting a genuine information problem, not just competing for visibility.

• Publish accurate state-specific procedure content. Correct classifications and penalties, dated and attorney-reviewed, so the accurate version is available for extraction.
• Cover the cooperation question carefully. Why voluntary interviews are rarely as voluntary as they seem, without overstating or fearmongering.
• Check what tools say about your state. Part of prompt auditing is verifying accuracy, because it shapes what your prospects believe before they call.
• Address misconceptions at intake. If prospects arrive with wrong assumptions from AI research, your intake should be prepared to correct them.
• Do not use AI to draft legal content unreviewed. The same failure modes apply to your own content production. Attorney review before publication is not optional here.
• Date and re-review everything. Statutes change and stale published penalties become the wrong answer these tools repeat.

4Why These Tools Hesitate to Name Attorneys

Ask for a criminal defense recommendation and you will often get a careful non-answer directing you to the state bar directory and advising you to consult several attorneys. Specific firm names appear less readily than in most local service categories.

That is deliberate caution around legal recommendations, and it tells you where the leverage sits. If these systems point people toward bar directories and verification sources, your completeness in those sources matters more than content volume alone.

• Official verification is the pathway. Bar directories are what these tools recommend. Your listing should be complete and accurate.
• Being cited as a content source is a win too. You may be named as the source of an answer about a charge even when no firm is recommended.
• Location-specific queries name firms more often. "Criminal defense attorneys in [county]" produces named results more readily than a general request.
• Specialization queries help. "Federal criminal defense attorneys in [district]" narrows the field and surfaces specific firms more often.
• Directories dominate here as in search. Avvo, FindLaw, Justia, and Martindale are structured and verifiable, which is what these systems prefer.
• Set expectations accordingly. Legal is a cautious category. Measure against that baseline rather than home services benchmarks.

5Crawler Access

• Check robots.txt for the major agents. GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and Applebot-Extended. Blanket disallows catch all of them.
• Check security plugins and firewalls. Aggressive defaults blocking unfamiliar user agents are the most common cause on law firm sites.
• Check host and CDN level blocking. Some blocking happens at the edge regardless of robots.txt.
• Google-Extended is separate from search indexing. Blocking it affects AI features only.
• Recheck after site changes. Rebuilds and plugin updates reintroduce blocks routinely.
• Decide deliberately. For a firm that wants to be found, permitting these crawlers is almost always correct.

6Attorney Entity Building for a Portfolio Practice

A portfolio practice has more entity work to do than a single-focus firm, because the tools need to understand not just who you are but what range of matters you actually handle.

1. State practice areas explicitly and completely. Every charge category you handle, named in plain language rather than implied through page structure.
2. Make bar admissions verifiable and consistent. Jurisdictions, dates, and bar numbers matching the state bar's public record exactly.
3. List federal court admissions separately. District and circuit admissions are distinct credentials that matter for federal matters and are strong entity anchors.
4. Build full bios per attorney. Including which practice areas each attorney handles, since a portfolio firm may have specialists within it.
5. Include board certification where held. Some states certify criminal law specialists, and it is verifiable and distinctive.
6. Name your counties and courts. Concretely, so location queries can match you.
7. Keep firm details identical everywhere. Website, Business Profile, bar listing, and every legal directory.

7Sources These Tools Trust

• State bar directories. The most authoritative source about an attorney and exactly what these tools direct people toward.
• Federal court admission records. Public and verifiable, and particularly relevant for federal practice claims.
• Criminal defense bar associations. State associations and the National Association of Criminal Defense Lawyers carry real weight in this practice area.
• Board certification bodies. Where your state certifies criminal law specialists, that record is authoritative.
• Legal directories. Avvo, Justia, FindLaw, and Martindale-Hubbell, claimed and accurate.
• Google Business Profile. Increasingly referenced for local business facts, which means local SEO work feeds AI visibility.
• Law school and CLE affiliations. Adjunct positions, alumni listings, and presenter credits are institutional verification.
• Local and legal press. Being quoted on a case or a law change is third-party authority no directory provides.

8Content Structured for Extraction

• Question headings with immediate answers. Phrase headings as a defendant would ask, then answer in the first sentence.
• State-specific facts throughout. Actual statute sections, classifications, and penalty ranges. Specificity is what gets extracted.
• Cover the pre-charge questions thoroughly. Target letters, detective contact, search warrants, and grand jury subpoenas, each as a substantial page.
• Add visible FAQ sections with matching schema. The questions you field on intake, answered plainly.
• Byline to admitted attorneys. With credentials, on all substantive legal content.
• Date and schedule re-review. Especially penalty content, since amendments make published figures wrong.
• Add appropriate schema. Attorney and LegalService markup, Person markup with credentials, FAQPage where FAQs are visible.

9Prompt Audits and Tracking

• Build a prompt set across your portfolio. Pre-charge questions, charge classification questions, penalty questions, and recommendation requests for each major practice area.
• Run monthly across the major tools. ChatGPT, Perplexity, Gemini, and Google AI Overviews. Results differ substantially.
• Verify accuracy, not just presence. Check whether what these tools say about your state's law is correct, because it shapes what prospects believe.
• Record who gets named. Your firm, competitors, directories, or nobody.
• Note the cited sources. They reveal where these tools draw from and where to invest.
• Test by charge category. Visibility may be strong for one practice area and absent for another.
• Expect variance. Responses are not deterministic. Track trends across repeated runs.

10Measuring AI Visibility

• Citation rate across the prompt set. Tracked month over month against a realistic baseline for legal.
• Pre-charge query citation specifically. The segment worth the most, tracked separately from charge queries.
• Referral traffic from AI tools. Modest volume, high intent, visible in analytics.
• Branded search growth. When a tool names you without linking, people search your name.
• Ask callers directly. An AI option in your intake question about how they found you catches what analytics cannot.
• Prospect sophistication at intake. Callers who already understand their charge classification and the risks of talking to investigators have researched somewhere.
• Keep expectations proportionate. This is a position being established rather than a channel producing volume today.