The SEC Marketing Rule and AI content: what advisors can and can't automate

Educational, not legal advice — your compliance officer and counsel own these calls for your firm.

Here’s the answer up front: the SEC Marketing Rule doesn’t prohibit AI-drafted content — it makes your firm responsible for whatever publishes, no matter what drafted it. That distinction tells you exactly what you can automate (drafting, pre-screening, archiving, distribution) and what you can’t (review, approval, substantiation, and the judgment calls the rule reserves for humans). Most advisory firms get this backwards: they treat AI as forbidden and publish nothing, while the real risk was never the tool — it’s unsupervised output.

What the rule actually cares about

The Marketing Rule — Rule 206(4)-1, in force for all RIAs since late 2022 — replaced decades of patchwork with principles. In plain English, an advertisement can’t: state something untrue or omit what’s needed to keep it from misleading; make a material claim you can’t substantiate on demand; imply something misleading; cherry-pick benefits without fair treatment of risks; or present performance in the ways the rule specifically restricts. Testimonials and endorsements are allowed but come with disclosure, oversight, and (in some cases) written-agreement requirements.

Notice what’s absent: any mention of how the content gets written. The rule is technology-neutral. A misleading sentence is a violation whether a partner, a freelancer, or a language model typed it — and a truthful, substantiated, reviewed one is fine on the same logic.

What you can automate

Drafting. The blank-page problem is real — most firms publish almost nothing because nobody has time to write. AI drafting solves the volume problem. The draft is the input to your compliance process, not the thing that goes live.

Pre-screening. This is the underused one. Before a human ever reads a draft, automated screening can catch the known landmines: performance promises, guarantee language, unhedged forward-looking statements, superlatives nobody can substantiate, testimonial content missing its disclosures. Your reviewer then reads clean drafts instead of rewriting risky ones — the review step gets faster and stricter.

Archiving. Books-and-records (Rule 204-2) requires retaining your advertisements and the support behind material claims. Automating the archive — every published piece, its approval trail, its substantiation — turns an exam scramble into a folder that already exists.

Distribution. Scheduling, posting, and sending approved content is mechanical. Automate freely — after approval.

What stays human

Review and approval. Someone at your firm reads every piece before it publishes and owns that decision. No screen replaces this; the screen exists to make it efficient.

Substantiation. If a piece makes a material statement of fact, a human confirms the firm can back it up — the rule expects a reasonable basis you can produce if asked.

Performance content. The rule’s performance provisions (net-of-fees presentation, prescribed time periods, restrictions on hypotheticals) are precise and unforgiving. Our own policy is blunt: automated drafting stays away from performance claims entirely. If your firm publishes performance, humans and counsel build it.

Testimonials and endorsements. Eligibility, required disclosures, compensation arrangements, oversight — judgment calls with per-case facts. Not automatable.

The workflow that survives an exam

The firms doing this well run the same loop: guardrails → draft → screen → human review → publish → archive. Define what content may never say (your policies, encoded). Draft with AI inside those guardrails. Screen automatically for landmines. Review by a human with authority to kill anything. Archive everything with its approval trail. Every step documented — because “we supervise our marketing” is a claim you’ll want to substantiate too.

That loop is how publishing more becomes compatible with publishing safely — which matters, because the engines answering your prospects’ questions (“fee-only fiduciary near me,” “is this firm legit”) cite firms that publish real answers. Silence isn’t a compliance strategy; it just cedes the answer to whoever figured this out first. That’s the argument we make at length on our financial advisors page.

See what the engines say about your firm

The free AI Visibility Report, advisor edition runs the diligence and recommendation queries your prospects actually ask — your firm’s name, your niche, your region — across every major AI engine, with screenshots. Built by hand, delivered in 48 hours, and useful to hand your CCO either way.