Recently, the AI community has been abuzz with discussions following a massive wave of account bans implemented by Claude. According to a report released by Anthropic in July 2026, Claude banned a staggering 11.4 million accounts in the first half of that year. Out of 398,000 appeals filed during the same period, only 42,000 were successful, marking a mere 10.5% success rate and highlighting the limited effectiveness of the appeals process. The reasons for these bans were varied, ranging from breaches of sensitive content red lines—such as illegal abuse—to crackdowns on accounts involved in black-market activities and usage in unsupported regions. The risk control mechanisms predominantly adopted a one-strike policy, where any violation resulted in a permanent ban. This led to many frequent users and developers losing their core work assets overnight due to account suspensions. Furthermore, the appeals process was riddled with issues, with numerous users unable to pinpoint the exact reasons for their bans. In some instances, the appeals portal even vanished subsequently. Currently, Anthropic has revised its usage policies, with new restrictions slated to take effect on November 12, 2026. Many users fear that another wave of large-scale account bans may ensue. These events serve as a stark reminder to practitioners not to rely solely on a single AI platform for their productivity assets.