Discord 已承认其 AI 安全系统存在一个重大错误，该错误自五月以来导致约 8,200 个账户被错误封禁。问题源于系统将电子表格和游戏纹理等无害图像误解为儿童性虐待材料。虽然通常由人工审核员审查标记内容，但一个错误导致了即时封禁，而另一个错误阻止了这些封禁被自动解除。Discord 此后已修复该问题，并正在努力实施更好的安全措施以防止类似事件发生。 AI
影响 凸显了过度依赖 AI 进行内容审核的风险，以及对强大的人工监督和更好安全措施的需求。
排序理由 一个广泛使用的平台审核工具中的错误导致了广泛的用户影响。
在 Mastodon — sigmoid.social 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
A bug may have led to around 8,200 erroneous bans on Discord.
Discord has admitted that a bug in its AI-powered content moderation system mistakenly banned over 8,000 users since May. The automated tool incorrectly flagged harmless images including spreadsheets, chessboards and game textures as harmful content. The company has now restored …
Discord has acknowledged that a bug in its AI moderation system mistakenly banned more than 8,000 users over the past two months, after harmless images—including spreadsheets, chessboards, game textures, as well as white and gray transparent backgrounds—were incorrectly flagged a…
Discord admits AI moderation bug wrongfully banned users over harmless images https://techcrunch.com/2026/07/07/discord-admits-ai-moderation-bug-wrongfully-banned-users-over-harmless-images/ # AI # Tech # SocialMedia