Home · Guides · Study

Florida tax-preparer email security, 2026

A public-DNS review of 7,661 Florida tax-preparer domains that have mail servers found that 82.1% do not enforce DMARC — the control that tells receiving mail systems to quarantine or reject forged messages using the firm’s domain. Results are published only in the aggregate. No firm is named. This is not a penetration test, not legal advice, and not an IRS or FTC publication.

Published 18 November 2026 by PreparerGuard / Noble Professional Services LLC, Windermere, Florida. Lookups were performed the same day.

Why this matters for tax offices

Criminals do not need a firm’s password to send a message that looks like it came from that firm. If the domain’s public DNS records do not tell receiving servers what to do with forged mail, those messages can land in a client’s inbox. The IRS and its Security Summit partners have repeatedly warned tax professionals about phishing, spear phishing, clone phishing (a near-copy of a legitimate message pretending to be the original sender), and “new client” lures. See, for example, IRS news release IR-2026-85 (4 August 2026).

SPF, DKIM, and DMARC are the public authentication records that make spoofing a firm’s domain harder. They are not a complete security program. Paid preparers also have a legal duty to maintain a written information security plan (WISP): Form W-12 (Rev. October 2026), line 11, and IRS Publication 5708 discuss that duty; the FTC Safeguards Rule at 16 CFR part 314 is the underlying framework. This study measures only the public email-authentication records. It does not measure whether a firm has a WISP, MFA, backups, or training.

Headline findings (mail-enabled domains, n = 7,661)

DMARC policy observedDomainsShare
No DMARC record4,02952.6%
Monitor only (p=none)2,24829.3%
Invalid DMARC record60.1%
Quarantine (p=quarantine)82610.8%
Reject (p=reject)5497.2%
Not enforcing (none + missing + invalid)6,28382.1%

A p=none record is useful for monitoring, but it does not ask receivers to hold back or refuse forged mail. We therefore count monitor-only and invalid records together with missing records as “not enforcing.”

SPF observationDomainsShare
No SPF record at the domain root7349.6%
More than one SPF record (invalid in practice)1121.5%
SPF ending in +all (explicitly permits anyone)00%

By apparent mail provider

Provider labels are inferred from public MX hostnames only (for example, Outlook/Microsoft 365, Google Workspace, or GoDaddy/Titan). “Other” includes every remaining MX pattern. A firm can use a provider and still leave DMARC off; the gap is usually a DNS setting, not a missing mailbox product.

Apparent providerDomainsNot enforcing DMARC
Microsoft 3652,92275.6%
Google Workspace1,96590.7%
GoDaddy/Titan16987.6%
Other2,60582.4%

By credential listed on the source file

Credential buckets come from the IRS FOIA PTIN extract (CPA, enrolled agent, attorney, or no such credential listed). A domain can appear in more than one bucket when several listed professionals share it, so the bucket sizes add up to more than 7,661 and are not a partition of the sample. Attorneys’ domains in this extract were the most likely to enforce DMARC; non-credentialed preparers’ domains were the least. That is a description of the aggregate, not a judgment of any person or firm.

Credential bucketDomains in bucketNot enforcing DMARC
Attorney32857.0%
CPA2,50077.6%
Enrolled agent1,08879.8%
No CPA / EA / attorney credential listed4,14383.2%

Methodology

  1. Starting list: the Florida extract of the IRS FOIA PTIN holder file, updated 14 August 2026 (68,766 rows). The file is a public records release of paid preparers; we used it only as a source of websites, not as a mailing list, and we did not contact the listed individuals or firms for this study.
  2. From listed websites we derived registrable domains, dropped free and social hosts (for example consumer webmail and social-network profiles), and de-duplicated. That left 9,482 unique domains.
  3. On 18 November 2026 we looked up each domain over DNS-over-HTTPS: whether it resolved, whether it published MX, the root TXT records used for SPF, and the _dmarc TXT record. 8,462 domains resolved; 7,661 had at least one MX record (the mail-enabled set reported above). 187 domains (2% of the unique set) returned lookup errors and were excluded from the percentages.
  4. We did not connect to any firm’s mail or web servers, did not send email, did not attempt logins, and did not test DKIM signatures (DKIM alignment was not measured). We do not publish, and we did not retain for publication, any per-firm result.
  5. “Not enforcing” means no DMARC record, a policy of p=none, or a record we could not parse as valid DMARC (six records). Quarantine and reject are counted as enforcing, even if other settings (rua, aspf, subdomain policy) may still be incomplete.

Limits — please read these before citing the numbers

What a tax office can usually do in a short sitting

  1. Look up the domain you actually send from (the one in your message From: header). A free check that runs in your browser, against public DNS only, is on our email spoofing guide and on the home-page checker.
  2. Publish a single SPF TXT record that includes the services you really send through, ending in ~all (or -all once you are sure the include list is complete). Remove duplicate SPF records. Do not use +all.
  3. Publish a DMARC record at _dmarc.yourdomain. A cautious start is p=none with an aggregate report address you will actually read. Then move to p=quarantine and, when mail still flows, p=reject.
  4. Record the change and the remaining work in your WISP so the written plan matches the office you run. See our Publication 5708 / Form W-12 line 11 guide.

Those steps are ordinary DNS and documentation work. They are not a substitute for legal advice, and they do not by themselves satisfy the Safeguards Rule.

Check your own domain. The in-browser checker reads public MX, SPF, DMARC, and common DKIM selectors over DNS-over-HTTPS. It does not log into your mail or store the domain you type. If you also need a written plan, PreparerGuard is a paid WISP kit we sell — that is a separate product, and this study does not depend on a purchase.

Check my email domain

Who paid for this, and what we are not claiming

This study was designed, run, and published by PreparerGuard, a product of Noble Professional Services LLC (Windermere, Florida). We sell WISP templates and a related domain-watch service to tax preparers. We have an interest in firms taking email authentication and written security plans seriously. We did not sell, and we will not sell, a list of the underlying firms. We do not claim IRS or FTC affiliation. Nothing on this page is legal advice, a security certification, or a finding about any named practice.

Sources and further reading

Cite as: PreparerGuard / Noble Professional Services LLC, “Florida tax-preparer email security, 2026,” 18 November 2026, https://preparerguard.com/guides/florida-tax-preparer-email-security-2026.