September 14, 2026 · Admin
Is Cold Email Legal in the US? Yes, Under Seven Federal Rules, and the Real Exposure Is State Court
Yes. Cold email is legal in the US, including B2B cold email, and federal law requires no opt-in of any kind. CAN-SPAM imposes seven rules and a $53,088 maximum penalty per email. The FTC is not who sues you.
That last sentence is where most guidance on this question stops being useful. Almost every page ranking for “is cold email legal in the US” recites the FTC’s seven requirements, quotes the penalty, and ends. The federal penalty is rarely applied to cold outreach. The live risk in 2026 is a state court class action over a subject line, brought by a private plaintiff who could not sue you under federal law at all. This piece sits in our outbound compliance hub and covers both layers. Every statute, case and figure is linked and dated at the end.
Is cold email legal in the US without permission?
Yes. CAN-SPAM is an opt-out statute, not an opt-in one. Nothing in 15 U.S.C. 7701 and following requires a prior relationship or any form of consent before you send a commercial email to a stranger. The law regulates how you send and what you must include, not whether you were invited.
It also draws no line between B2B and B2C. The statute applies to any “commercial electronic mail message,” defined at 15 U.S.C. 7702(2)(A) as a message “the primary purpose of which is the commercial advertisement or promotion of a commercial product or service.” A prospecting email to a VP of Operations at a manufacturer is squarely inside that definition. The absence of a B2B carve-out is the reason so much US outbound exists at all, and it is the single biggest difference between the US regime and the consent-based regimes in Europe and Canada.
One exception matters operationally. A “transactional or relationship message,” defined at 16 CFR 316.3, is exempt from most of the requirements below: messages that facilitate an already-agreed transaction, deliver warranty or safety information, notify about account changes, or relate to employment. A cold prospecting email is never one of these, and dressing one up as a transactional notice turns a compliance question into a deception question.
What are the seven rules CAN-SPAM actually imposes?
Seven, and all of them are free to comply with. The FTC publishes them in plain language in its own compliance guide for business, the only source on this topic worth treating as authoritative. Here they are, with what each means for a cold email specifically.
| Rule | What the FTC requires | What it means for cold email |
|---|---|---|
| 1. Accurate header information | “Don’t use false or misleading header information.” From, To, Reply-To and routing data must be accurate. | Sending domains must resolve to a real business. Alias domains are fine. Forged ones are not. |
| 2. Honest subject lines | “Don’t use deceptive subject lines.” The subject must reflect the content. | The rule that generates almost all the litigation. See the state law section below. |
| 3. Identify the message as an ad | Disclose “clearly and conspicuously” that the message is an advertisement. | The FTC allows latitude in how. A signature identifying the company and the offer generally does it. |
| 4. Include a physical address | A valid physical postal address. A registered PO box or private mailbox is acceptable. | The most skipped requirement in cold email, and the cheapest to fix. |
| 5. Provide an opt-out | A “clear and conspicuous explanation” of how to stop receiving email. | A plain “reply STOP” satisfies this. A one-click link is safer. |
| 6. Honor opt-outs within 10 business days | Process requests within 10 business days and keep the mechanism working at least 30 days after sending. | Suppression must be global across every sending domain and inbox you operate. |
| 7. Monitor anyone sending for you | “Monitor what others are doing on your behalf.” Hiring a vendor does not transfer legal responsibility. | The rule that decides agency liability. Its own section below. |
One requirement inside rule 5 is easy to breach without noticing. 16 CFR 316.5 bars charging a fee, requiring any information beyond the recipient’s email address and opt-out preference, or making the recipient do anything more than reply or visit a single web page. An unsubscribe flow that asks for a login, a reason, or a survey answer is a violation on its own.
What is the penalty, and who can actually sue you?
The FTC’s published maximum is $53,088 per email, adjusted annually for inflation. Individual recipients cannot sue you under CAN-SPAM at all. That combination is why the federal statute looks terrifying on a compliance slide and almost never produces a bill.
Enforcement authority under 15 U.S.C. 7706 sits with the FTC, other federal agencies, state attorneys general, and providers of “internet access service.” That last category is the only private right of action in the statute, and the Ninth Circuit narrowed it close to nothing in Gordon v. Virtumundo, 575 F.3d 1040 (9th Cir. 2009): a plaintiff must qualify as an internet access service and show actual adverse effect from the specific messages. Running your own mail server does not get you there.
For scale: the largest CAN-SPAM settlement the FTC has announced remains the $2.9 million ValueClick case of 17 March 2008, and the conduct there was deceptive “free laptop” offers and unencrypted payment data, not B2B prospecting. That is the reassuring half. Here is the half that has changed.
If not the FTC, where does the risk actually come from?
State court, from private plaintiffs, over subject lines. Two statutes carry nearly all of it: California Business and Professions Code 17529.5 and Washington’s Commercial Electronic Mail Act, RCW 19.190. Both give the recipient a direct right to sue with per-email statutory damages, exactly what CAN-SPAM withholds.
| Regime | Who can sue you | Damages per email | Cap |
|---|---|---|---|
| CAN-SPAM, 15 U.S.C. 7706 | FTC, other federal agencies, state AGs, internet access services | Up to $53,088 | None stated; court discretion |
| California, B&P 17529.5 | The recipient, an email service provider, or the Attorney General | $1,000 liquidated, or $100 if you maintained reasonable preventive practices | $1,000,000 per incident, or $100,000 with reasonable practices |
| Washington CEMA, RCW 19.190, filed before 11 June 2026 | The recipient | $500 statutory, strict liability | None stated |
| Washington CEMA, filed on or after 11 June 2026 | The recipient | $100 statutory, knowledge required | None stated |
The California statute makes it unlawful to send a commercial email containing a third party’s domain name without permission, falsified or forged header information, or a subject line “that a person knows would be likely to mislead a recipient” about a material fact. Note the shape of it: $1,000 per email, class-wide, plus fees, awarded to the recipient. A ten thousand message campaign with one arguably misleading subject line is a theoretical $1,000,000 incident. Defense counsel report a rising volume of demand letters and class filings under this section, and the trigger list they describe includes deceptive subject lines, unauthorized domain references, missing opt-out mechanisms, and failures of SPF, DKIM and DMARC authentication. Authentication is not a CAN-SPAM requirement. It is turning up in complaints anyway, which is reason enough to treat it as one.
Washington moved faster. On 17 April 2025 the Washington Supreme Court decided Brown v. Old Navy, LLC, holding that RCW 19.190.020(1)(b) bars false or misleading subject lines and that whether one is true “does not depend on what the rest of the email conveys.” A subject line reading “Last chance” for a sale extended the following week became actionable on its own. Class filings followed at $500 per message under strict liability, and the legislature responded with HB 2274, effective 11 June 2026, cutting damages to $100 and adding a knowledge requirement. Cases filed before that date keep the old standard.
Does CAN-SPAM preempt state email laws?
Mostly, and not in the part that matters. 15 U.S.C. 7707(b)(1) supersedes any state statute “that expressly regulates the use of electronic mail to send commercial messages,” but carves out any such law “to the extent that [it] prohibits falsity or deception in any portion of a commercial electronic mail message.”
Both California 17529.5 and the Washington provision are drafted to sit inside that carve-out, so they survive. On 14 January 2026 the US District Court for the Western District of Washington rejected a CAN-SPAM preemption defense to CEMA in Ma v. Nike, Inc., No. C25-1235JLR, finding Congress had preserved state authority over misleading subject lines. Federal compliance is a floor, not a shield: you can satisfy all seven FTC rules and still lose a California class action over a subject line.
Who is legally the sender when an agency sends for you?
You are. Under 15 U.S.C. 7702(16)(A) the “sender” is the person who initiates the message and “whose product, service, or Internet web site is advertised or promoted by the message.” That is the client, not the agency. Hiring someone to press send does not move the definition.
The agency is not off the hook either. “Initiate” at 15 U.S.C. 7702(9) includes procuring the transmission of a message, which captures the vendor, and the FTC’s seventh rule is explicit that the advertiser cannot contract the obligation away: “Monitor what others are doing on your behalf.” Both parties can be liable, and the one whose logo is in the signature is the one the statute names first.
Nobody in the category writes about this, and it should drive how you paper an outbound engagement. Three things belong in the contract before you sign:
- Who owns the sending domains. If the agency owns them and the relationship ends badly, your suppression list and domain reputation leave with them. If you own them, a deliverability incident follows you. Decide deliberately rather than by default.
- Where the data came from. Ask for list provenance in writing. “Scraped” and “bought from a broker” are different answers with different exposure, particularly if any recipient is in the EU.
- Indemnification, both directions. Whoever writes the subject lines carries the subject line risk; whoever supplies the list carries the list. Most agency contracts are silent here, which means the client absorbs everything.
Weigh this alongside what the engagement costs. A cheaper retainer with an uncapped indemnity gap is not cheaper.
What does a compliant B2B cold email look like in practice?
Seven checks, none of which cost money. Run the list against a template once and it holds for every message the sequence sends.
- The subject line is literally true. No fake “Re:” or “Fwd:”. No “as discussed” when nothing was. No urgency claim you will contradict next week. After Brown, the subject line is judged on its own, not against the body.
- The From name and domain identify a real business a recipient can look up in ten seconds.
- A physical postal address is in the footer. A PO box or registered private mailbox is acceptable to the FTC.
- The opt-out is visible and free. One reply or one click, no login, no form fields beyond email address, no reason required.
- Opt-outs suppress globally within 10 business days, across every domain and inbox, not just the one that got the reply.
- Authentication passes. SPF, DKIM and DMARC aligned on every sending domain. Not required by CAN-SPAM, showing up in state complaints, and what decides whether any of this reaches an inbox. The failure modes are in why cold emails go to spam.
- Someone owns the suppression list as a named job. Most opt-out failures are operational, not legal.
Call a lawyer if you are sending at volume into California or Washington, if any recipient is in the EU, UK or Canada, or if you have received a demand letter. The first two are jurisdictional questions this page cannot answer for your list. The third is not a research problem.
How LeadButton handles this
Every campaign we run ships with the seven FTC requirements built into the template before the first send: real sending domains registered to a real entity, a postal address in the footer, a one-step opt-out, and suppression that is global across every inbox and domain in the program rather than per-sequence. Subject lines are written to be literally true, which after Brown v. Old Navy is a legal position and not a style preference.
Launch is $1,500 a month for managed email outreach, Growth is $3,500 a month for email plus LinkedIn, and Scale is custom. Domain ownership, list provenance and the indemnity split come up in the first call rather than staying buried, because the statute names you as the sender and you should know what you are signing. Pricing is on the LeadButton pricing page. We are not lawyers and give no legal advice; what we can tell you is which of the above we handle as a matter of course and which sit with you.
Sources
- Federal Trade Commission, CAN-SPAM Act: A Compliance Guide for Business, checked 14 September 2026. Source for the seven requirements, the $53,088 per email maximum penalty, the 10-business-day opt-out rule and the 30-day mechanism rule.
- 16 CFR Part 316, CAN-SPAM Rule, eCFR current version, checked 14 September 2026. Source for the primary purpose test and the prohibition on fees or extra information in opt-out mechanisms.
- 15 U.S.C. 7702, Definitions, Cornell Legal Information Institute, checked 14 September 2026. Source for the definitions of commercial electronic mail message, initiate and sender.
- 15 U.S.C. 7707(b)(1), Effect on other laws, Cornell Legal Information Institute, checked 14 September 2026. Source for the preemption clause and the falsity or deception carve-out.
- Cornell Legal Information Institute, CAN-SPAM Act: Private Right of Action for Internet Access Service Providers, checked 14 September 2026. Source for who may sue and for the Gordon v. Virtumundo, 575 F.3d 1040 (9th Cir. 2009) status and injury requirements.
- California Business and Professions Code 17529.5, Justia, checked 14 September 2026. Source for the prohibited conduct, the $1,000 and $100 liquidated damages tiers, the $1,000,000 and $100,000 per-incident caps, standing and attorney’s fees.
- Arnold & Porter, Washington State’s CEMA Amendment: A Speedbump, Not a Roadblock, published 30 June 2026, checked 14 September 2026. Source for HB 2274, the 11 June 2026 effective date, and the change from $500 strict liability to $100 with a knowledge requirement.
- Clark Hill, Washington Supreme Court on false or misleading subject lines, published 18 April 2025, checked 14 September 2026. Source for Brown v. Old Navy, LLC, decided 17 April 2025, and the holding that subject line truthfulness does not depend on the rest of the email.
- Faegre Drinker, Federal Court Upholds Washington State’s CEMA Against CAN-SPAM Preemption Argument, published 22 January 2026, checked 14 September 2026. Source for Ma v. Nike, Inc., No. C25-1235JLR (W.D. Wash. 14 January 2026).
- Freeman Mathis & Gary, California’s Anti-Spam Law Takes Center Stage in a Growing Wave of Class Action Lawsuits, published 25 March 2026, checked 14 September 2026. Source for the categories of conduct alleged in current 17529.5 filings, including authentication failures.
- Federal Trade Commission, ValueClick to Pay $2.9 Million to Settle FTC Charges, 17 March 2008, checked 14 September 2026. Described by the FTC as the largest settlement in a case based on the CAN-SPAM Act.
[…] is smaller than that figure implies, and the mechanics of who counts as the sender are covered in whether cold email is legal in the US. The opt-out rule itself is not the part anyone argues […]
[…] opt-out law, so you do not need permission before a first cold email. The email itself falls under the federal CAN-SPAM rules and California’s anti-spam statute, which apply whether or not you meet the CCPA […]