Adrian Alexandru Stinga

The Criminal groups Does the Intrusion. The State linked groups Only Has to Shop.

Part of GN-090 Report which is available to read for free on Aether-Intel

The cybercriminal selling stolen credentials for $500 thinks they're making another routine marketplace sale.

What if the real...

Inside the sHUMINT Methodology: Part IV — Attribution of AI attacks

Everyone wants to know which AI model was used.

That's useful but it's only the starting point.

Models don't choose objectives. People do.

In Part IV of the sHUMINT methodology, I explore why attribution should move...

Ransomware isn't just cybercrime anymore. Not because every operator has geopolitical ambitions. But because thousands of independent profit-driven decisions can collectively produce the strategic effects a hostile state wants. From the perspective of many ransomware...

The third article in my Synthetic HUMINT (sHUMINT) methodology series is now live.

Part III: Consistency Testing of AI explores why contradictions are more valuable than agreement when analyzing AI behavior. Rather than treating inconsistencies as simple errors, I argue they can...