Checklist: how to screen an AI project for an empty shell in 15 minutes

A sector where hype outruns product by multiples requires a filter. Here is ours - five questions answerable with open data, 15 minutes of time.

The checklist

• 1. Who pays? Look for revenue: utilization dashboards, on-chain fees, public client contracts. No answer - no business.
• 2. What does the token do? Payments, staking to run nodes, a share of fees - fine. "Governance" and "ecosystem incentives" are a polite "nothing".
• 3. Is there a product without the token? A working demo, an API, users. If the entire product is the token and a roadmap, you are looking at a meme.
• 4. Who is on the team and what is behind it? AI engineers with a track record - or anons with a template website? Serial token launchers are a red flag of their own (we track those in the deployers section).
• 5. Distribution and unlocks. Insider share, the vesting schedule, who is selling into the book. Tokenomics kills more often than a bad product.

A project that passes all five is a rarity, and that is normal: in a young sector most projects are shells. Our reviews of specific AI tokens with verdicts - in the section.