Era 08 · 2020s — Now

AI Trading

Lag

Lag forming now

Pulse cycle

0.3s

AI systems are trading trillions. Large language models are reading news feeds and adjusting positions in milliseconds. No specific federal framework governs this class of behavior. The largest regulatory gap in financial history is being created in real time — and it is on track to be the longest-lived.

Events in this era

2020s

ML-Driven HFT Dominance

By the mid-2020s, algorithmic and high-frequency trading firms handled roughly 60 to 70 percent of US equity volume. Many now rely on machine-learning models retrained on intraday data. Their systems adjust quotes in microseconds based on patterns human analysts cannot articulate. Firms like Citadel Securities, Virtu, and Jane Street intermediate a large share of retail orders. The SEC's Reg SCI governs exchange infrastructure. The models themselves sit inside private trading firms, where direct regulatory inspection is limited.

~60–70% of US equity volume is algorithmic

2025 — 2026

No Framework

As of this writing, there is no US federal regulatory framework specifically governing AI-driven trading models, their training data, or their behavior in stressed markets. The EU's AI Act is the closest analog, but carves out most financial applications.

Largest active regulatory gap in financial history

Aug 1, 2024

EU AI Act enters force

The first comprehensive AI regulation from any major jurisdiction. Classifies AI systems by risk level with increasing compliance requirements. High-risk applications require transparency, auditing, and human oversight. Phased implementation runs through 2026.

First systematic framework; carve-outs for most trading

← Previous era

Crypto

Loading signal…