Can agents already run a campaign end to end? What goals and guardrails must already be in place? Is the fully autonomous version coming soon? This panel puts the question of agentic trading to people who draw the line in different places. Which decisions are agents more capable than humans at, and can you trust agents to do this yet? Are agents able to chase the outcomes you care about? What must be true about your data and your oversight before you'd let an agent off the leash? And when it gets it wrong, who owns that?