OpenLedger launches a human-centered AI agent, prioritizing safety over speed

While most other crypto AI tools focus on automated trading speed, San Francisco-based OpenLedger has taken a different approach. They have introduced a new AI agent, designed to put control back in the hands of users, not executing any trades without their explicit approval.

Safe, transparent, and traceable

The standout feature of #AIAgent is the "Proof of Attribution" system. Every trading decision the AI proposes is directly linked to specific data sources and versions of the AI model, allowing users to fully verify and understand the reasoning behind each action. If the data chain is incomplete or untraceable, the AI will reject the trade, helping to prevent risks from inaccurate data.

According to Kamesh, a core member of OpenLedger, their goal is to make AI a "smart co-pilot" for users, rather than a "black box." Users can set risk limits and review each proposal before execution. This AI agent will be integrated into #TrustWallet and begin testing with a small group of users in October, before expanding to the market early next year.

OpenLedger's approach is a solution to concerns about AI's "hallucinations," where system errors can lead to significant financial damage. It emphasizes the need for transparency and human involvement in the era of automated trading.