When a server detects a high request rate, a suspicious traffic pattern, or repeated policy violations from a given IP address, it may add that IP to a blocklist. Future requests from the banned IP receive a 403 Forbidden, a TCP connection reset, or are silently dropped. Bans can be temporary (minutes to hours, often tied to 429 rate-limit responses) or permanent (until manually reviewed).
IP bans are the most common defence against scraping at scale. They are also the primary reason why proxy rotation is essential for high-volume scraping: spreading requests across thousands of IP addresses means any single ban removes only a small fraction of total capacity, and the banned IP can be retired from the pool.
Residential and mobile proxies are harder to ban than datacenter proxies because their IPs are shared with real users — banning them would block legitimate traffic. Anti-ban strategies also include reducing request rates, adding realistic delays between requests, and rotating user agents and session cookies alongside IP addresses.