◉ PSYCHOHISTORY

RECTor (Robust Efficient Correlation Attack on Tor, 2025-2026)

mechanismDarknet & Cyber
A new AI technique makes it far easier to unmask people who thought Tor kept them anonymous.
Who they are

RECTor, a 2025-2026 academic method for attacking the Tor anonymity network.

What they do

It uses machine learning to dramatically improve the efficiency of timing attacks that match up traffic to identify users.

How it works

By applying contrastive learning, it can deanonymize traffic flows even when an attacker only captures part of the packets or faces heavy background noise.

Why it matters

It represents the cutting edge of the attacking side in the ongoing cat-and-mouse fight over Tor anonymity.

The engine's record — word for word
2025/2026 academic framework using machine learning + contrastive learning to vastly improve Tor traffic-correlation timing-attack efficiency. Allows adversaries to deanonymize flows even with partial packet capture or high background noise. Operational state-of-the-art for the offensive side of the perpetual Tor arms race.
Follow the trail
Walk this on the live map →
Part of the Psychohistory engine — 2,437 entities, 6,337 documented connections. Open data, built to be proven wrong.