Workload throttling is the deliberate reduction of service capacity, request rate, or processing speed after a usage limit, risk threshold, or policy rule is reached. It is common in cloud platforms, APIs, and AI services, where providers may slow responses, queue jobs, or cap concurrent tasks instead of fully blocking access.
In cyber security, throttling matters because it can protect systems from abuse and resource exhaustion, but it can also become an operational bottleneck. Defenders use it to limit brute-force logins, stop scraping, and keep shared infrastructure stable during spikes. Attackers may trigger throttling on purpose to disrupt monitoring, delay backups, or exhaust a victim’s quota-based services. In practice, teams need to track limits, distinguish soft throttling from hard outages, and design fallback paths for long-running workloads.