Short answer: yes, on both. Substack's rules don't mention AI at all. beehiiv allows clearly labelled AI content, but bans a newsletter that is entirely AI-written with no real human input, and anything that makes readers believe a human wrote it when no one did. The safest format is the one most guides miss: you write it, as yourself, about her.
Substack: no AI rule. Its Content Guidelines (last updated 29 September 2026) ban impersonation, plagiarism, spam and publications that are mainly marketing. Nothing about AI-made writing or characters.
beehiiv: an AI section with two lines. Its Acceptable Use Policy (last modified 16 July 2026) says labelled AI content "may be permitted" if it doesn't mislead readers "into believing the Content is human-authored when it is not". And: "Publications that rely entirely on AI-generated material without meaningful human input are not permitted."
Both ban bought or scraped lists. Only people who opted in to your publication.
Both ban a newsletter that is just an ad. Substack: no publications "whose primary purpose is to advertise external products". beehiiv: occasional affiliate links are fine, an affiliate-first newsletter isn't.
Fees. Substack keeps 10% of paid subscriptions plus Stripe's fees. beehiiv keeps 0%, but paid subscriptions need a paid plan.
Read on 3 November 2026 on Substack's and beehiiv's own pages (sources at the end). This is a reading of public policy pages, not legal advice.
A newsletter is the one channel your followers can't lose to an algorithm.
Yes. Neither Printful nor Printify bans AI-made designs. Both write AI-generated content into their IP policies: you promise you have the rights to every design, nothing may break anyone's copyright, trademark or publicity rights, and if you're unsure you must not upload it. Their own AI tools add that you can't pass the output off as human-made or as a real photo. Both start free; the paid plans cost $24.99 a month.
November 3, 2026 · 9 min read
Every other channel an AI influencer lives on can change its mind about her. Instagram adds a label, TikTok changes the payout bar, an account gets flagged. A newsletter list is the one audience you take with you. It's also a second way to earn: paid subscriptions, sponsor slots and, on beehiiv, an ad network.
For a character account it does one more thing: it's the natural place to show how she's made. People who follow an AI influencer are often curious about exactly that.
Substack's rules, read for an AI character
Substack has no AI-content rule. What matters for a character account:
Impersonation. "We don't allow impersonation, which includes posing as another person, brand, or organization." A fictional character isn't another person. A character built to look like a real one is. The Terms add that you can't pick another person's name "with the intent to impersonate that person".
Plagiarism. "Do not publish any material that was written or created by someone else and claim it as your own." That rule is about other people's work; it's a good reason to keep a character's posts original rather than rewritten from someone else's newsletter.
Marketing. Substack is "not conventional email marketing". A newsletter that exists to push affiliate links or drive traffic is out; brands "may be subject to additional verification".
Lists. Imports must be people "who explicitly opted-in to receive emails from your specific publication". No bought, scraped or harvested addresses, no importing your contacts.
Synthetic images. No sexual images of a person without consent, "including synthetic or AI-generated ones", and no sexualised depiction of anyone who appears to be a minor, "regardless of whether the content is synthetic". Erotic writing is allowed; profile pictures must have no nudity.
Age. You must be at least 16 to use Substack at all.
One detail in Substack's Terms (updated 6 October 2026) is worth knowing: they mention "AI-detection features" and "the application or removal of any AI-detection label", and say Substack makes no promise that a detection result is accurate. The Terms don't say when a label goes on. The simple answer is to label her yourself, in the publication's About page and in the first line of the welcome email, so a detector never tells your readers something you didn't.
On beehiiv the question isn't whether AI helped. It's whether a human did too.
beehiiv's rules: the human-input line
beehiiv is stricter, and it says so in a section called "Use of AI". It "supports the responsible use of AI tools", but a newsletter must stay "authentic, editorially driven, and valuable to readers". It names what's banned:
a publication that relies "entirely on AI-generated material without meaningful human input";
AI used "to mass-produce repetitive material", like templated listicles or a generic thought of the day;
several newsletters with slightly reworded AI content;
AI used to "impersonate individuals, or otherwise deceive readers".
And in the section on manipulated media: clearly labelled AI content can be fine as long as readers aren't led to believe it was written by a human. The policy also bans "Misrepresenting authorship".
For an AI influencer that means two things. First, label her: a newsletter signed by a character, with no word that she's AI, is exactly what the policy describes. Second, there has to be a person doing real editorial work: choosing what goes out, checking it, adding something. A pipeline that writes and sends issues on its own, with nobody reading them, is the case beehiiv rules out.
beehiiv's money features have their own rules:
Ad Network. Ads come with an automatic "Sponsored" or "Ad" label you can't hide. No clicking your own ads, and no newsletter "created primarily to display ads".
Recommendations (paid sign-ups between newsletters). No bot or fake sign-ups, and no "thin, AI-generated, or templated publications" set up to collect payouts. Breaking these can cost you every monetization program at once.
Affiliate links. "Occasional use of affiliate links is allowed"; a newsletter made mainly for affiliate income isn't.
The two rules side by side
Question
Substack
beehiiv
AI-written content allowed?
No AI rule in the Content Guidelines
Yes, if labelled and not passed off as human-written
Fully automated newsletter?
No specific rule
Not allowed without "meaningful human input"
A fictional AI character?
Fine; impersonating a real person isn't
Fine if labelled; impersonating individuals isn't
Bought or scraped lists
Banned
Banned
Newsletter that's mainly ads or affiliate links
Banned
Banned (occasional affiliate links are fine)
Cut of paid subscriptions
10%, plus Stripe
0%, plus Stripe; paid plan needed
Free plan
Free at any size
Free up to 2,500 subscribers
What a $5 subscription leaves you
Substack's fee page lists: 10% of each subscription payment, Stripe card processing of 2.9% + $0.30 for US cards, and Stripe Billing of 0.7% for recurring payments. On a $5 monthly subscription paid by a US card:
Substack 10%: $0.50
Stripe 2.9% + $0.30: about $0.45
Stripe Billing 0.7%: about $0.04
You keep about $4.02.
beehiiv's pricing FAQ says it takes "0% of your paid subscription" revenue, "minus Stripe's standard processing fee of 2.9% + $0.30". That leaves about $4.56 of the same $5. But paid subscriptions start on the Lite plan; when we read the page, it showed $49 a month with yearly billing.
So the choice is mostly about size. Substack's 10% equals beehiiv's $49 a month at about $490 a month in subscriptions. Below that, Substack costs you less; above it, beehiiv does. (Plan prices change with your subscriber count; check the slider on the pricing page for your number.)
Be honest with yourself about the "below" part. Most new newsletters, like most new accounts, start with a handful of readers. A free newsletter that grows the list is usually the right first step.
Only people who asked for it. Both platforms ban bought lists.
Write it as yourself, about her
There are two ways to run an AI influencer's newsletter:
In her voice. The character "writes" each issue. This works on Substack and, labelled, on beehiiv, as long as a person really edits it. It's also where the risk sits: one unlabelled issue and readers feel fooled.
In your voice, about her. You, the person behind the account, write a short behind-the-scenes letter: what she posted this week, which ideas got cut, what the tools cost, what went wrong. That is plainly human-authored, so beehiiv's human-input line is met by design, and it's the kind of thing her followers are curious about.
We'd start with the second. In our own pipeline, the rejected frames are often more interesting than the kept ones: in our first 672 critic reviews, only 48% of frames were kept, and 18% of the rejected ones were turned down for looking AI-made. A weekly "what we cut and why" is honest, easy to write, and hard to fake. We do the same thing in public with our weekly critic digest.
A setup that stays inside the lines
Pick the platform by size. Free list: either one. Paid subscriptions under about $490 a month: Substack costs less. Above it: beehiiv.
Say she's AI where readers sign up. The About page, the welcome email, and the signature of every issue. Our free AI disclosure checker writes a disclosure line you can adapt.
Keep a human in the loop. Read every issue before it goes out. On beehiiv it's a rule; everywhere else it's how you avoid sending something wrong to everyone at once.
Grow the list honestly. A sign-up link in her bio and posts; no bought lists, no importing contacts, no fake sign-ups for Recommendations.
Disclose sponsors and affiliate links in the issue itself. On beehiiv's Ad Network the label is automatic; your own deals need your own label.
Keep her face her own. A character that looks like a real person is the impersonation both platforms ban.
Does Substack ban AI-generated newsletters? No. Its Content Guidelines have no AI clause. The Terms mention AI-detection features and "any AI-detection label", and make no promise that a detection result is accurate. Label her yourself.
Can my AI influencer sign the newsletter in her own name? On Substack, nothing forbids it, as long as she isn't impersonating a real person. On beehiiv, only if readers can see she's AI: the policy bans leading them to believe a human wrote it.
Can I let AI write and send every issue automatically? Not on beehiiv: publications that "rely entirely on AI-generated material without meaningful human input are not permitted". Substack has no such rule, but an unread, auto-sent issue is how mistakes reach your whole list.
Which is cheaper for a small paid newsletter? Substack, until you reach roughly $490 a month in subscriptions; then beehiiv's 0% cut beats the plan fee. Both add Stripe's card fees.
For every platform's rules in one place, see the short answers on /rules.
Make yours
Influgen gives every AI influencer a small team: a head writer for the ideas, a creative director and stylist for each scene, and a critic who rejects frames before they post. It makes short posts for Instagram, TikTok and X, labelled as AI. It doesn't write or send newsletters; the list, and the letter you write about her, are yours. Start here, or try the free tools first.
Sources
Substack, Content Guidelines (last updated 29 September 2026; read 3 November 2026)
Substack, Terms of Use (last updated 6 October 2026; content moderation, AI-detection and age sections; read 3 November 2026)
beehiiv, Acceptable Use Policy (last modified 16 July 2026; "AI-Generated and Manipulated Media", "Use of AI", Ad Network and Recommendations sections; read 3 November 2026)
beehiiv, Pricing (plans and FAQ; read 3 November 2026)
Yes. Shopify's Acceptable Use Policy has no AI rule and Stripe lets creators sell their own content from their own site without pre-approval. The limits are about what you sell and what you claim: Stripe bans adult content (AI-made included), IP-infringing goods, get-rich-quick offers and fake testimonials, and the FTC's review rule bans an avatar inventing a customer experience. Shopify Basic costs $39 a month ($29 yearly); cards cost 2.9% + 30¢.
Yes. Twitch's Community Guidelines don't mention AI at all, and the one place they name VTubers applies the same content rules to avatars as to people. What they ban is deepfakes used to abuse or sexualise a real person, impersonation and faked viewer counts. To earn, the channel needs Affiliate: 25 followers, 4 hours streamed on 4 different days and 3 average viewers. Subscriptions split 50/50.
