ShanHaiYin

Detection Insights

Guides, tests, and case notes about AI content detection, with an emphasis on verifiable evidence and practical methods.

Latest articles

We publish practical detection tests, methodology notes, and verified public references. Articles involving current events identify their sources and publication dates.

DETECTION EXPLAINER · SEPTEMBER 1, 2026

What Does AI Image Detection Actually Look For?

It is not just about weird fingers or strange eyes. A practical detector checks provenance, file metadata, editing traces, pixel patterns, and context.

Read article →
FAMILY SAFETY GUIDE · AUGUST 14, 2026

Deepfake Doctor and Supplement Scams: How Families Can Verify AI Health Videos

Miracle-cure videos and supplement ads may use AI-generated doctors, fake celebrity endorsements, and synthetic patient stories. Preserve the media, verify outside the ad, and use reports as supporting evidence.

Read article →
AUDIO & VIDEO VERIFICATION · AUGUST 13, 2026

That Voice May Not Be Real

AI voice cloning and deepfake video scams can impersonate executives, family members, and trusted contacts. Preserve the file, verify outside the original channel, and use technical reports as an early warning layer.

Read article →
EVIDENCE VERIFICATION · AUGUST 9, 2026

Are On-Site Photos Real? Verifying Contractor, Rental, Damage, and Claim Images

How to preserve original photos or videos, check suspicious files, and request stronger follow-up evidence before payment, reimbursement, refunds, or sign-off.

Read article →
SELLER RISK GUIDE · AUGUST 8, 2026

Can a Payment Screenshot Be Fake? A Seller’s 3-Step Verification Checklist

How sellers should verify PayPal, Venmo, Cash App, Zelle, bank-transfer, and card-payment screenshots before releasing goods or refunds.

Read article →
REGULATORY GUIDE · AUGUST 4, 2026

How to Understand China’s Measures for Labeling AI-Generated and Synthetic Content

A practical guide to explicit labels, file metadata, platform checks, user declarations, and the role of GB 45438—2025.

Read article →
DETECTION PRINCIPLES · AUGUST 4, 2026

How Does AI Image Detection Work? From Provenance Credentials to Pixel Review

How ShanHaiYin fingerprints a file, validates C2PA, reads standardized AIGC metadata, and uses model and pixel analysis only when needed.

Read article →
VERIFICATION GUIDE · AUGUST 3, 2026

How to Tell if an Image Is AI-Generated: A 7-Step Verification Guide

Preserve the original file, trace earlier versions, inspect Content Credentials and platform signals, then weigh visual and model evidence.

Read article →
DETECTION METHODS · 2026-08-01

Is AI Text Detection Reliable? False Positives, Limitations, and Best Practices

Why short passages, mixed human-AI writing, editing, translation, and unfamiliar styles can affect a result—and how to use detection responsibly.

Read article →
Detection methods · July 23, 2026

Why Are Cropped AI Images Harder to Detect?

How provenance credentials, embedded signals, and pixel-level features can change between an original image, screenshot, and crop.

Read article →