@agt_ai

AI Guard Protocol (AGT) is a BSC-based project claiming to offer AI-powered fraud detection tools including deepfake scanning and smart contract auditing.

The project has a token ($AGT, 500M supply) and appears to be post-TGE with airdrop distribution completed.

However, the project exhibits several concerning patterns: presale-heavy tokenomics (10% presale + daily distribution schemes), withdrawal restriction announcements suggesting custody issues, and generic AI claims without verifiable technical substance.

Green flags: Working product claims with multiple AI tools (image/video/voice deepfake detection, contract scanner) · Completed airdrop distribution and active development updates · Clear tokenomics documentation with locked allocations

Red flags: Presale-heavy distribution model with complex daily reward schemes typical of yield farming projects · Multiple withdrawal-related announcements suggesting liquidity or custody problems · Generic AI security claims without technical validation, open-source code, or third-party audits · BSC-only deployment with CEX listing fund allocation (classic pump setup) · Tier B scout with modest 55/100 track record · Suspicious timeline: account created Feb 2026 but tweets reference July 2026 (future dates suggest data issues or fabrication)