What changed
HYPOTHESIS (from convergence description, no source signals attached): Meta reportedly flipped Instagram photos to AI-training/generation material by default, allowing anyone to generate AI images of a public account's likeness via an @-mention prompt without per-use consent. NOTE: the input provided zero signals and zero citations, so even the core factual premise is unverified in this brief and must be confirmed before any build.
Why now
If true, the default flip is days old and small-business awareness is near zero; every future Meta default/policy change re-opens exposure, which is the only thing that converts a one-time scare into a recurring subscription. This 'why now' is entirely inference — no demand_evidence was supplied.
Converging signals
Only one underlying signal (1571, creator/small-business brand protection) is referenced and its text was not provided. This is a single-signal pattern instantiation ('Passivity Penalty Flip'), not a true multi-signal convergence. Convergence strength is weak.
Customer pain
HYPOTHESIS: boutique brands, realtors, coaches, and creators whose face IS the brand do not know their likeness is now generatable by strangers, and would feel violated once shown proof. There is NO provided evidence (no complaints, no threads, no hiring/spend data) that anyone currently feels or pays for this pain. The pain must be manufactured via demonstration, which is fear-marketing — churn-prone and reputationally delicate.
Who pays
HYPOTHESIS: solo personal brands with revenue attached to their face (realtors, coaches, boutique agencies, OnlyFans/creator managers, personal-brand consultants). Reachable via small-business Facebook groups, realtor associations, creator communities. Willingness to pay is unproven; adjacent proof (people pay for brand-protection/DMCA services like Red Points, BrandShield at much higher price points, and creators pay takedown services) is from general knowledge, not provided sources — treat as inference.
Solved today
Mostly not solved: owners either don't know, manually dig through Instagram settings once, or hire social-media managers who may toggle the setting. Reputation-monitoring tools (Mention, Brand24) watch text mentions, not likeness generatability. No provided evidence of an incumbent doing proof-of-exposure probing.
Why current solutions are bad
One-time manual opt-out decays: Meta's rolling default changes across surfaces (Instagram, Facebook, WhatsApp, Meta AI) silently re-open exposure, and nobody re-verifies. No tool produces PROOF (an actual generated image of the client) which is the only artifact that makes the risk visceral enough to sell.
Proposed product
A monitor that (a) with client consent, periodically attempts generation of the client's likeness via the @-mention flow and archives the result as proof-of-exposure, (b) watches Meta policy/default changes and alerts when exposure re-opens, (c) walks through or executes the opt-out/settings fix and re-verifies it worked, with a monthly 'exposure report' PDF the client can also show their own customers.
MVP version
No-code-heavy MVP in 2-3 weeks: manual/semi-automated probing from a test account against 30 consenting pilot accounts, a Meta-policy-change watcher (scrape newsroom/help-center diffs), a Loom-style proof video per client, Stripe subscription, landing page. The convergence's own testable prediction IS the MVP test: probe 30 consenting accounts, launch a $19/mo waitlist page in 2-3 communities, require ≥3% waitlist conversion and majority generatability to proceed.
30-day build
Week 1: verify the factual premise from primary sources (Meta help docs, newsroom, reputable tech press) — if the @-mention generation flow or the default flip is not confirmed, kill immediately. Weeks 2-3: recruit 30 consenting pilot accounts, run probes manually, document generatability rate. Week 4: landing page + waitlist promoted in 2-3 realtor/coach communities; measure conversion.
60-day build
If ≥3% waitlist conversion: build the recurring probe + policy-watch + opt-out-verify loop (Playwright-driven where ToS-safe, manual-assisted where not), onboard first 20 paying subscribers at founder pricing ($19/mo or $190/yr), produce the monthly proof report automatically.
90-day revenue plan
Target 50-150 subscribers at $19-49/mo ($1k-5k MRR) via community posts, realtor-association newsletters, and a free one-time 'exposure check' as lead magnet. Also sell a $99 one-time 'audit + lockdown' for non-subscribers as cash-flow wedge.
Distribution path
Free instant 'can a stranger generate YOUR face?' check as the hook; demonstrated-value sales (show them their own AI-generated image) matches the founder's demo-not-relationship style. Channels: realtor Facebook groups, coach/creator communities, local business associations, short-form video of the probe itself.
Pricing hypothesis
$99 one-time audit+lockdown; $19/mo solo monitor; $49/mo multi-account (agencies managing several personal brands). Low enough for card-swipe, no procurement.
Technical difficulty
Low-to-moderate build (scraping, browser automation, Stripe, report generation — all founder strengths). The hard part is not code: it is that probing must be done from accounts Meta could ban, and reliable automated probing at scale against Meta surfaces is an adversarial cat-and-mouse game (see lesson: even Reddit blocks datacenter IPs; Meta is far more aggressive).
Legal / regulatory risk
Moderate: generating images of clients' likenesses is consensual (contracted), but automated probing almost certainly violates Meta's automation/ToS rules — the business's core loop depends on tolerated ToS violation. Also fear-based marketing around likeness rights invites scrutiny. Not heavily regulated, but the ToS exposure is structural, not incidental.
Platform dependency
EXTREME — this is the kill-shot dimension. The product exists only in the gap Meta leaves open: Meta shipping a prominent one-tap permanent opt-out (their likely PR response to backlash), blocking @-mention generation of non-followers, or banning probe accounts each independently destroys the product. The convergence's own falsification criteria admit this.
Founder fit
Mixed. Fits: complaint-adjacent niche monitoring tool, micro-SaaS, fast AI-assisted prototyping, demonstrated-value selling, no VC needed. Does not fit: this is NOT the founder's proven government-portal/forced-buyer shape (lesson, conf 0.80) — there is no mandate, no deadline, no forced filer; the 'registry' is a private platform's mutable settings page. It also brushes against his stated avoidance of consumer-ish, platform-policy-risk products. Founder COULD build it in days; the question is whether he should.
Breakout potential
Modest: could expand to TikTok/X/YouTube likeness policies, deepfake monitoring, and state likeness-rights laws (e.g. right-of-publicity compliance for agencies), becoming a general 'likeness rights ops' tool. But each expansion inherits the same platform-adversarial posture.
Final recommendation
PARK / CHEAP-TEST ONLY. Do not commit build time yet. The idea is testable for ~$50 and one week (verify premise from primary sources, manually probe consenting accounts, waitlist page). Proceed past that gate ONLY if: premise confirms, majority generatability holds, waitlist converts ≥3%, AND Meta has not announced a permanent opt-out. Even then, treat as a cash-flow side product with a 6-12 month expected lifespan, not a durable business — it fails the founder's best pattern (forced buyer with statutory deadline) and lives entirely inside a hostile platform's tolerance.
Next action
Spend one day verifying the factual premise from Meta's own help-center/newsroom pages and two reputable tech-press sources; if confirmed, recruit 10-30 consenting accounts from one realtor Facebook group and manually run the @-mention probe, archiving results.