ChatGPT can be made to generate sexualised and violent images, researchers find
Published: 2026-06-18 13:58:56 Author: Editor 
Comments
Comments来源：BBC News · 作者：Editorial · Wed, 17 Ju· 分类：GlobalBut researchers say its still possible to trick th
。
来源：BBC News · 作者：Editorial · · 分类：GlobalBut researchers say its still possible to trick the AI chatbot into producing graphic content.原文链接：https://www.bbc.com/news/articles/c802ldjdklzo?at_medium=RSS&at_campaign=rss Related Articles
Google has ended Nest Mini and Nest Audio production, discontinuing both speakers
Google has confirmed that it has ended production of the Nest Mini and Nest Audio, its 5+ year old s2026-06-18
1Password Acquires Apono in Reported $250M-$300M Deal
来源：SecurityWeek · 作者：Eduard Kovacs · Wed, 17 Ju· 分类：CybersecurityApono specializes in just-in-time a2026-06-18
Adversarial Exposure Validation Turns Security Visibility into Confident Prioritization
来源：The Hacker News · 作者：[email protected] (The Hacker News) · Wed, 17 Ju· 分类：CybersecurityFor s2026-06-18
Smashing Security podcast #472: AI gets hacked, and BitLocker gets bypassed
来源：Graham Cluley · 作者：Graham Cluley · Wed, 17 Ju· 分类：CybersecurityWhat if your AI coding assistant c2026-06-18
Parnell Square trial: Doctor on scene describes knife injuries inflicted on child
Riad Bouchaker (52), of no fixed address, is on trial at the Central Criminal Court charged with the2026-06-18
Google, Microsoft, and OpenAI Back Linux Foundation's Appia AI Standards Initiative
来源：Slashdot · 作者：BeauHD · 2026-06-17· 分类：TechnologyBrianFagioli writes: Google, Microsoft, OpenAI, A2026-06-18
Latest Comments