Swarm Intelligence Under Adversarial Inputs and Hard Time Budgets: A Comparative Review of Drone-Based OSINT Verification and High-Frequency Trading

We review swarm intelligence (SI) in two demanding domains: drone swarms that gather physical evidence to corroborate open-source intelligence (OSINT), and high-frequency trading (HFT). We argue that both are instances of decentralized evidence aggregation under adversarial inputs and hard time budgets, and compare them through that lens. In drone OSINT, SI's value lies in coverage, task allocation, consensus, and fault tolerance, while verification itself remains a human judgment. In HFT, algorithms such as particle swarm and ant colony optimization are used offline for tuning and portfolio construction rather than in the latency-critical trading loop. We discuss scalability, security against adversarial agents, evaluation practice, ethics, and regulation, and pose open research questions.