Staying authentic and legal
This lesson exists because two clocks are ticking. Audiences are getting sharper at detecting machine-made content by the week, and regulators are writing rules by the quarter — and a one-person brand can't afford to learn either lesson the expensive way. The good news: staying on the right side of both is mostly one policy applied consistently, and you've been building it since lesson 1. Here it is, assembled.
What the rules actually ask of you
The regulatory picture as of mid-2026, in plain terms, converges on one word: disclosure. Platforms increasingly label AI-generated and AI-altered media — Meta marks such content both by detection and self-declaration — and the direction of travel everywhere is that realistic synthetic media should say so. India's advertising watchdog ASCI, meanwhile, has long required that paid partnerships and material connections be declared, and its guidance has been extending into AI-era territory: the safe reading for a small brand is that deception is the offence — an AI-generated "photo" passing as real, a synthetic testimonial, an influencer likeness that never said those words. Details will keep shifting (and vary by where your audience is), but notice what never shifts: rules against pretending. Your compliance strategy is simply not pretending — which you already adopted in lesson 5 for images, and which costs nothing when practised from day one.
Copyright runs the other direction and surprises everyone: the question isn't only "can I use AI images?" but "can I own them?" In most jurisdictions today, purely machine-generated work sits in murky-to-absent copyright territory — India's courts and registrars, like most, are still working it out — meaning the AI mood-image you posted may be unprotectable against a competitor lifting it. One more quiet argument for the rule this course keeps landing on: your photographs of your products, your words in your voice, are unambiguously yours — legally, and in every other sense.
The audience clock is the stricter one
Whatever regulators settle, your followers rule on authenticity daily, and their standard is stricter and faster. The pattern in audience research and platform behaviour is consistent: people accept AI assistance (nobody resents a well-drafted caption) and punish AI impersonation — fake photos, hollow at-scale personalisation, the uncanny sameness of unedited model output (the slop from lesson 1). The line your audience actually polices is the line between AI helped me say this and a machine pretended to be me. So the durable policy, one paragraph, into the voice doc:
Machines may draft anything; only true things ship. Real products and people are photographed, not generated; imagined visuals are labelled. Facts get checked by the human whose name is on the account. Words leave in our voice, after our edit. When asked whether we use AI: yes, gladly, for the boring parts.
That last line deserves its own beat, because founders fear the question. The strong answer isn't denial — it's Meera's, cheerfully: "AI drafts my captions and sorts my DMs so I can spend mornings in the workshop; every photo is real and every fact is mine." Audiences don't resent leverage. They resent being fooled — and a brand that's relaxed about its tools reads as having nothing to hide, because it doesn't.
Do this today: paste the policy paragraph into your voice doc, edited to your reality. Then audit your last month against it — any image that could be mistaken for a real product shot, any shipped fact you never verified. Fix forward; nobody's grading the past.