ScamTime
An AI persona app for small businesses that routes suspicious emails, texts, and DMs to a friendly AI character who plays along with the scammer, wastes their time, and produces an evidence report you can hand to your bank or local fraud unit.
Solo founders and small shop owners drowning in phishing texts, fake invoice scams, and bogus supplier DMs
- One-click 'send to scammer' — AI persona replies in character for up to 20 messages without revealing real business data
- Auto-extracted fraud report listing scammer wallet addresses, callback numbers, and bank details
- Daily digest showing attempt volume, scam category mix, and risk score trend across the week
- Shareable evidence pack formatted for local police fraud reports and bank reimbursement claims
LLM-powered honeypots moved from research curiosity to genuinely usable in 2025-2026, and small businesses now report scam losses exceeding shoplifting losses — yet no consumer-grade tool exists for the owner who handles their own inbox.
FTC reported $3.5B in imposter scam losses in 2025 (nearly tripled since 2020) and published an April 2025 Data Spotlight specifically on text scams targeting businesses; BBB maintains an entire Business Scam Study catalog.FTC warns of record $3.5 billion losses to imposter scams in 2025 ↗New FTC Data Spotlight highlights text scams that may target your business ↗2025 Business Scam Study - Better Business Bureau ↗
No consumer-grade, multi-channel (email/text/DM) AI scam-bait product surfaced; the closest is an open-source developer project (ScamShield) on GitHub, while enterprise honeypots (T-Pot, Cowrie, Thinkst Canary) are aimed at security teams, not solo founders.ScamShield: An agentic AI-powered honeypot (GitHub) ↗Best Honeypot Solutions in 2025 (including free options!) ↗
No public pricing exists for a consumer AI scam-bait product; solo founders are notoriously price-sensitive and LLM cost-per-engagement can be non-trivial, but adjacent small-biz security tools (email filtering, scam detection) do monetize via SaaS subscriptions.Top 10 AI Scam Detection Tools in 2026: Features, Pros, Cons & Comparison ↗Phishing Statistics 2026: Latest Trends & Risks ↗
Phishing volumes projected at 5M+ attacks/year, global phishing sites crossed 1M in 2025, and AI is making scams more convincing — the problem is structurally worsening, not cyclical.Phishing statistics 2025 - 2026: The numbers you need to know ↗Phishing Statistics 2026: Losses, AI Attacks & Attack Volume Data ↗
ScamShield on GitHub proves the core LLM-persona honeypot loop already works in production-ish form; remaining work is multi-channel routing (email/SMS/DM) + report packaging, all standard SaaS engineering.ScamShield: An agentic AI-powered honeypot (GitHub) ↗