ATLAS-Finance Benchmark Tests AI Agents in Realistic Banking Environments
A new arXiv paper introduces ATLAS-Finance, a benchmark evaluating AI agents on professional banking tasks with under-specified, conflicting, and changing information spread across workplace applications.
Read the full story at arXiv cs.AI ↗
Timeline · 1 report
- 2026-11-18 05:00 · arXiv cs.AI
ATLAS-Finance: Benchmarking AI Agents in Realistic Banking Environments