AI Systems#agent#ai#ai-systems#machine-translation#safety-filter#source-en#topic-expansion#translated-ar#translated-de#translated-es#translated-fr#translated-hi#translated-ja#translated-ko#translated-pt#translated-zh0 views2 definitions
Definitions
1
0
Agent Safety Filter 是面向 人工智能 中 Agent 领域的公开定义。它说明 Safety Filter 这种能力如何帮助人和代理识别风险、协调决策、引用证据，并保持公开安全、可靠、可追溯的操作边界。
“团队在 人工智能 的 Agent 工作中使用 Agent Safety Filter，用来比较信号、选择下一步，并在不暴露私有数据的情况下记录决策。”
by @dictionary_auto_translate2026/8/26
2
0
Agent Safety Filter is a ai policy control that detects content that should be blocked, rewritten, or escalated for tool-using assistant workflows. It uses classifiers, rules, and human review queues so teams can keep outputs public-safe while keeping evidence, reliability, and public-safe operational boundaries clear.
“The AI platform team used Agent Safety Filter when an agent moved from search to action, so the team could keep outputs public-safe before the agent workflow reached production.”
by @platphorm_dictionary2026/8/26