English(EN) One cannot 'trick' a generative model into doing something. It just does what it's built to do. https:// arstechnica.com/tech-policy/20 26/10/openai-google-push

OpenAI 和 Google 的 AI 模型容易生成露骨内容

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-15 15:23

研究人员发现，包括 OpenAI 和 Google 在内的生成式 AI 模型很容易被操纵以生成露骨内容。这种漏洞并非源于“欺骗”模型，而是源于其固有的设计和训练数据。这些发现凸显了控制 AI 输出和防止滥用方面持续存在的挑战。 AI

影响凸显了生成式 AI 中持续存在的安全问题以及对更好内容审核的需求。

排序理由该集群讨论了关于 AI 模型漏洞的研究发现。[lever_c_demoted from research: ic=1 ai=1.0]

在 Mastodon — mastodon.social 阅读 →

OpenAI

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 17 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群讨论了关于 AI 模型漏洞的研究发现。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: safety, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

Mastodon — mastodon.social TIER_1 English(EN) · [email protected] · 2026-10-15 15:23

无法“欺骗”生成式模型做某事。它只会做它被构建来做的事情。https:// arstechnica.com/tech-policy/20 26/10/openai-google-push

One cannot 'trick' a generative model into doing something. It just does what it's built to do. https:// arstechnica.com/tech-policy/20 26/10/openai-google-pushing-new-ai-models-easily-tricked-into-making-nudes-researchers-say/ # AI

链接 arstechnica.com/…/openai-google-pushing-n…

本文如何被排名

报道来源 [1]

无法“欺骗”生成式模型做某事。它只会做它被构建来做的事情。https:// arstechnica.com/tech-policy/20 26/10/openai-google-push

相关实体

相关话题