Working at the frontier: How Balyasny Asset Management evaluates and governs Claude Fable 5
Balyasny Asset Management says 2026 marks the shift from AI that searches to AI that does work, using Anthropic’s Claude Fable 5 plus in-house agent infrastructure to automate complex financial analysis, improve speed and quality, and enforce strict governance through approved tools, human review, and enterprise controls.
MAIN POINTS
- BAM evaluates models on thousands of real financial tasks, not generic benchmarks, to measure practical performance and failure modes.
- Claude Fable 5 improved complex planning and agentic execution, outperforming the prior model on relevant tasks.
- BAM built BAMAgent and custom harnesses to securely run long, multi-step workflows with approved data, permissions, and review controls.
- Frontier AI now enables tasks like merger analysis, tax-loss harvesting, and macro research to finish in minutes instead of days.
TAKEAWAYS
- Enterprise AI value depends as much on workflow design and controls as on model capability.
- Real-world evaluations with verifiable outcomes can reveal meaningful model differences hidden by standard benchmarks.
- Stronger models do not replace governance; they increase the need for access limits, logging, and human oversight.
- BAM sees the next phase of AI as agents becoming teammates that proactively support investment work.