Autonomous AI Traded $20m in Crypto on Testnet
A study reveals that autonomous large language model agents successfully conducted $20 million worth of cryptocurrency trading with real capital on a restricted testnet over 21 days.
What happened?
In a study published on arXiv, researchers investigated the reliability of autonomous LLM agents converting user mandates into validated tool actions. It was conducted on DX Terminal Pro, a 21-day deployment where 3,505 user-funded agents traded real ETH in a bounded on-chain market. The system generated 7.5 million agent calls and approximately 300,000 on-chain actions during the period. Over 5,000 ETH was utilised during the experiment.
Key facts
| Handelsperiod | 21 dagar |
|---|---|
| Antal agenter | 3 505 |
| Total handelsvolym | $20 miljoner USD |
| Utplacerat ETH | >5 000 ETH |
| Antal onchain-åtgärder | Cirka 300 000 |
| Lyckade transaktioner | 99,9% |
”We study reliability in autonomous language-model agents that translate user mandates into validated tool actions under real capital.”
”The system produced 7.5M agent invocations, roughly 300K onchain actions, about $20M in volume, more than 5,000 ETH deployed, roughly 70B inference tokens, and 99.9% settlement success for policy-valid submitted transactions.”
”Reliability did not come from the base model alone; it e”
Why it matters
The study's results indicate that the system's reliability did not stem solely from the base model. This highlights the importance of external control systems for the safe operation of AI agents. The 99.9% success rate for approved transactions demonstrates the potential for AI agents in financial applications under controlled conditions. The experience also provides insights into how AI agents can handle financial transactions with real capital.
Who is affected?
This development impacts researchers and developers within AI, particularly those working with autonomous agents and decentralised applications (dApps). Financial institutions and cryptocurrency exchanges may also benefit from the insights. Users considering AI-driven financial tools may be interested in the study's conclusions regarding reliability and control.
What else you should know
The DX Terminal Pro experiment generated a comprehensive dataset. The system handled approximately 70 billion inference tokens. Long-running agents made thousands of sequential decisions, with the most active performing over 6,000 prompt-state-action cycles.
Quick answers about this story
Vad har hänt?
När hände det?
The link opens in a new window and leads to the publisher's own site.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.
AI-verktyg i artikeln
Topics
Get similar news straight to your inbox
The reader's room
Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.
Sign in to submit a comment or question.
Read the article through your role
- Decide whether this affects strategy over 6–12 months or is just noise.
- Discuss with leadership: do we own the right question or does ownership need to move?
- Ask: what risk are we taking by NOT acting on this this quarter?
Generated angle — not editorial analysis of "Autonomous AI Traded $20m in Crypto on Testnet"