Maryland Telemarketing Law for Insurance Agents (2026)

Maryland’s Stop the Spam Calls Act caps telemarketing calls at three per day to the same person, restricts calling hours to 8 a.m. to 8 p.m., and requires prior express written consent before you use an automated system, a recorded voice, or an artificial voice to reach a Maryland consumer, under Sections 14-4501 and 14-4502 of the Maryland Commercial Law Article. It’s been the law since January 1, 2024, and it applies to an insurance agent’s outbound follow-up call the same way it applies to a robocall scam, with no carve-out for “we’re just trying to reach our own leads faster.”

If you sell in Maryland, or you buy leads that include Maryland numbers, this is the rulebook your calling cadence has to fit inside, whether you dial by hand or run an AI caller. Get the hours or the frequency cap wrong and a single called party can sue you directly, on top of whatever the state’s consumer protection enforcement can add.

What Maryland’s Stop the Spam Calls Act actually says

The statute lives in the Maryland Commercial Law Article, Title 14, Subtitle 45, added by House Bill 37 in the 2023 legislative session and signed into law as Chapter 414, effective January 1, 2024, according to the Maryland General Assembly’s own bill record. It passed both chambers unanimously, 134 to 0 in the House and 46 to 0 in the Senate, which tells you how little political appetite there was to leave this alone.

Three rules do the actual work, all in Section 14-4502:

The calling window. “A person may not make or cause to be made a telephone solicitation… to a called party during the hours between 8 p.m. and 8 a.m. in the called party’s time zone.” That’s a full four hours tighter than the 8 a.m.–9 p.m. window most agents are used to from the federal Telemarketing Sales Rule. An agent who dials a Maryland lead at 8:20 p.m. out of habit, because that call would be fine in a state without its own restriction, is already outside the statute.

The frequency cap. No more than three calls to the same called party in a 24-hour period “on the same subject matter or issue, regardless of the telephone numbers used to make the call.” Read that last clause carefully. It doesn’t matter if call one came from your main line and calls two and three came from two different local-presence numbers; the cap tracks the person being called and the subject of the call, not which of your numbers dialed it.

The caller ID rule. You cannot fail to transmit your originating number, or intentionally act to prevent your name or number from reaching the called party, when your equipment is capable of sending it. A separate clause in the same section bars intentionally altering your voice to disguise your identity in order to defraud, confuse, or injure the called party, or to extract personal information under false pretenses. That second piece targets scam-style voice spoofing specifically, but the caller ID transmission requirement applies to every covered solicitation call, full stop.

None of this replaces the federal Telephone Consumer Protection Act. It sits on top of it. The TCPA, at 47 U.S.C. Section 227, separately requires prior express consent before you call a wireless number using an automatic telephone dialing system or an artificial or prerecorded voice, per Cornell Law School’s Legal Information Institute. Maryland’s statute adds a specific calling window, a specific daily cap, and its own private right of action that a called party can bring regardless of what they could also pursue under the TCPA. Clearing the federal bar doesn’t mean you’ve cleared Maryland’s.

What counts as “the same subject matter”

The frequency cap only bites on calls “on the same subject matter or issue,” which raises the obvious question: does a call about a new quote and a separate call about a renewal count against the same three-call limit, or do they run on independent counters? The statute itself doesn’t hand you a bright-line test, and neither Section 14-4502 nor its legislative history spells out a formula for what separates one subject matter from another. That ambiguity cuts against the caller, not in favor of them, because the safest reading, and the one a plaintiff’s attorney will push for, treats “we’re following up on your insurance” broadly enough that a quote call and a renewal call the same week look like the same subject matter to a called party who’s getting three calls from your agency regardless of which internal reason prompted each one.

Build your tracking around that conservative reading. Count every call from your agency to a given person toward the same three-call ceiling in a 24-hour window unless you have a clear, defensible reason to treat two topics as genuinely separate, such as an inbound service call the client initiated themselves versus an outbound sales solicitation you placed. A system that tracks “calls about Product A” and “calls about Product B” as separate three-call buckets to the same person is optimizing for a reading of the statute nobody has tested in Maryland court yet, and testing it is not a bet worth making with statutory damages on the other side.

Texts and ringless voicemail count too

Section 14-4501’s definition of prior express written consent doesn’t limit itself to phone calls. It explicitly covers a telephone solicitation “by telephone call, text message, or voicemail,” and separately names the playing of a recorded or artificial voice message and the transmission of a prerecorded voicemail as the specific activities that require consent. That means an automated SMS follow-up sequence and a ringless voicemail drop into a Maryland number both need the same written consent record as a live AI voice call, not a lighter-touch version of it because the medium feels less intrusive.

This matters because a lot of agencies treat text and voicemail drops as the “safe” channel precisely because they don’t ring the phone. Maryland’s statute doesn’t share that assumption. If you’re already checking whether ringless voicemail is legal for your leads more broadly, the mechanics of federal exposure are covered in our piece on number warmup and spam labeling; Maryland’s consent and frequency rules apply on top of that, specifically for Maryland numbers, the same as they do for a live call.

Why a state law like this exists on top of federal rules

Maryland isn’t alone in doing this, and it isn’t the first. Oklahoma, Florida, Texas, Georgia, Oregon, Virginia, and Pennsylvania have all layered their own telemarketing restrictions on top of the federal TCPA in recent years, each with its own combination of calling windows, consent standards, and private enforcement. The pattern is consistent enough that it has a name in compliance circles: a “mini-TCPA.” States pass them because federal enforcement is resource-constrained and slow relative to the volume of unwanted calls consumers report, and a state legislature can give its own residents a private right of action that doesn’t depend on the FCC or FTC picking their complaint out of a queue.

For an insurance agency, the practical effect is that “compliant” isn’t one national setting you configure once. It’s a set of overlapping, state-specific rules that all apply simultaneously to the same list the moment that list includes people in more than one state, which almost every agency’s does. A campaign built around the federal Telemarketing Sales Rule’s 8 a.m.–9 p.m. window and no explicit daily cap will violate Maryland’s narrower window and three-call limit on every Maryland number in the list, even though nothing about the campaign looks aggressive by federal standards alone.

What it actually costs to get this wrong

Two enforcement tracks run independently, and getting hit by one doesn’t protect you from the other.

Run the arithmetic on the private right of action alone. It’s a per-call, per-person exposure, not a one-time fine for the campaign. A list of 400 Maryland numbers dialed once outside the calling window, or dialed a fourth time in a day, is 400 separate potential $500-to-$1,500 claims (tripled if a court calls it willful), not one violation with one price tag. The statute’s non-waivable clause closes the obvious loophole too: you cannot bury a waiver of these rights in a sign-up form or a terms-of-service agreement unless it’s part of settling an actual legal dispute.

Put a number on the scenario that actually happens: an agency runs an end-of-month push on a 1,200-name multi-state list, and 90 of those numbers turn out to be Maryland numbers nobody flagged. If the campaign’s default calling window runs to 9 p.m. and ten of those ninety contacts get a call between 8 and 9 p.m., that’s ten separate statutory violations from one unnoticed scheduling gap, before anyone even looks at whether the frequency cap held. At the statute’s floor of $500 per claim, that’s $5,000 of exposure from a single evening’s dial session on a list most agencies would call clean. The private right of action doesn’t require the called party to prove they were harmed beyond the statutory minimum; it requires proving the call happened outside the rule, which is a matter of a timestamp and a call log.

Does this apply to an AI caller specifically?

Yes, and this is worth stating plainly rather than hoping the answer is different for automation than it is for a human dialer. Section 14-4502 covers “the playing of a recorded or artificial voice message when a connection is completed to a number called,” which is exactly the mechanism an AI voice caller uses. Separately, at the federal level, the FCC adopted a declaratory ruling on February 2, 2024, released February 8, 2024, confirming that the TCPA’s “artificial or prerecorded voice” prohibition covers AI technologies that generate human-sounding voices, closing off any argument that a synthetic voice is somehow outside the older statutory language written before that technology existed.

Put together, an AI caller reaching a Maryland number needs the same prior express written consent, the same 8 a.m.–8 p.m. window, and the same three-call cap as a human telemarketer running a manual dial session. Nothing about the call being automated, or sounding conversational instead of like a recorded announcement, moves it outside the statute. If anything, automation raises the stakes, because an AI caller can generate the volume to hit a frequency-cap violation across hundreds of numbers in an afternoon in a way a single human dialer physically can’t.

How to fix it: build the compliance into the campaign, not around it

None of this requires exotic tooling. It requires building your calling cadence around Maryland’s actual rules instead of your default settings, and you can do this yourself with a CRM, a spreadsheet, and some discipline before you ever touch an automated platform.

1. Flag Maryland numbers in your list before you build any campaign. Area codes 240, 301, 410, and 443 cover Maryland, though the safest method is the called party’s stated or verified address, not just the area code, since ported numbers travel with people across state lines. A list that isn’t tagged by state can’t be governed by state-specific rules.
2. Set your calling window to the narrowest rule that applies, not the loosest. If your list spans Maryland and looser states, run the whole batch inside Maryland’s 8 a.m.–8 p.m. window rather than maintaining two separate schedules that are easy to mix up under deadline pressure.
3. Cap attempts at three per 24 hours per person, system-wide, regardless of which number placed the call. Track this against the called party, not the outbound line, since the statute explicitly says the cap applies “regardless of the telephone numbers used to make the call.”
4. Collect and store prior express written consent before a Maryland number enters any automated campaign. The record needs the specific phone number and a disclosure that automated or artificial-voice calls may follow, per Section 14-4501’s own definition. A verbal “sure, go ahead” on an inbound call isn’t the same thing as a signed or e-signed written agreement.
5. Never block or suppress your caller ID on outbound Maryland calls. Number warmup and spam-label management are legitimate practices for keeping your caller ID from getting flagged as spam by carriers; deliberately failing to transmit your number, or spoofing a different one to conceal your identity, is the specific conduct Section 14-4502 prohibits.

The calling window, visualized

Maryland’s statute doesn’t leave the window ambiguous. Half the day is open for a covered solicitation call, and half is closed, with no exceptions carved out for insurance follow-up specifically.

How we solve it

TheAffordableAI runs outbound and inbound calling built to fit inside a calling window and a per-person frequency cap as system rules, not as a checklist a producer has to remember under deadline pressure. Number warmup and spam defense keep your caller ID transmitting cleanly, which is both a “Spam Likely” problem and, per Section 14-4502, a legal requirement in Maryland specifically. When a Maryland lead wants to talk it through, the call transfers warm to a licensed person on your team while they’re still on the line, and the call record, including consent notes, syncs back to your CRM, HighLevel included, automatically. The full rundown of number warmup, warm transfers, and CRM sync is on the features page.

What you get

The compliance part, said plainly

Speed is not a reason to skip the rules, and automating your calling doesn’t move the responsibility off you.

Where this isn’t the right tool

An AI caller enforces the mechanical rules well: the window, the cap, the consent gate, the caller ID. It cannot tell you whether a given lead’s consent record actually meets Maryland’s written-consent standard, and it shouldn’t be trusted to make that call on its own. If your Maryland book is small enough to track in a spreadsheet you actually look at every day, that may be all you need; the cost of a platform isn’t worth it for a few dozen names a month. And if a Maryland client calls with a coverage question that needs real judgment rather than a scheduling confirmation, that conversation belongs with a licensed person, not an automated script, regardless of how well the automation handles the calling-hours math.

It’s also worth saying plainly that no calling platform, ours included, can retroactively fix a consent record that was never properly collected. If your list includes Maryland numbers pulled from an aged database or a lead source you can’t document consent for, the fix is upstream: get consent or drop the number, not find a faster way to call it anyway.

Where to start this week

1. Tag every Maryland number in your active list. Area code alone isn’t reliable given number porting; cross-check against the called party’s stated or verified address where you can.
2. Set your outbound calling window to 8 a.m.–8 p.m. for Maryland contacts specifically, even if your platform’s default is wider. If your list is mixed-state, the simplest fix is running the narrowest rule across the whole batch.
3. Confirm your consent records actually meet Section 14-4501’s standard before any Maryland number goes into an automated campaign: a written or e-signed agreement naming the specific number and disclosing automated contact, not just a general opt-in checkbox from a lead form.

Maryland’s rules aren’t complicated once you write them down: an 8 a.m.–8 p.m. window, three calls a day per person, written consent before automating anything, and your caller ID transmitting every time. What’s hard is holding all four of those at once, by hand, across a list that includes Maryland numbers mixed in with everyone else’s, every single day, without a single slip. That’s the part worth automating. The law itself isn’t going anywhere.