OpenAI and Anthropic on Balancing Autonomy and Safety in Financial Agents
daniel_mac8 · x · 2026-10-15
Key theme at AI Engineer NY: balancing agent autonomy with safety
At the AI Engineer New York conference, frontier labs and financial firms discussed how to give agents useful autonomy while keeping consequential decisions under control. Main lesson: safe autonomy depends on where you place controls and human judgment, not how many restrictions you impose.
- OpenAI: Joseph Kim, Head of ChatGPT for Financial Services, introduced a "threshold of usefulness" — the point where a model becomes capable and reliable enough to actually save time. He predicted rising model capability will soon push financial analysts over this threshold, letting them delegate more work to agents, and that by 2027 the finance industry will rely less on Excel and PowerPoint.
- The article also covers Anthropic's approach and two finance firms' perspectives.
More from coding & agent
- Stop letting Claude or Codex pick your stack: write it in AGENTS.md and gate new dependencies — weswinder · 2026-10-15
- Dev closes all PRs: let agents patch open source via Nix overlays instead — kieranklaassen · 2026-10-15
- Prime Sandboxes Adds Filesystem Checkpoints to Fork Agent Rollouts Mid-Trajectory — samsja19 · 2026-10-15
- Every to host behind-the-scenes session on reporting Codex with Codex — every · 2026-10-15
- x402 transactions on Base jump 65% in three quarters, Chainalysis says — kleffew94 · 2026-10-15
- Pydantic AI author boosts PiDrei, a Python port of Pi using Rust-based Monty — samuelcolvin · 2026-10-15