OpenAI's Dots and Meta's Muse agent privacy promises under scrutiny
What happened
The Verge reported on October 10 that OpenAI's Dots and Meta's Muse both pitch privacy and safety to users, but differ in their privacy promises, data controls and problems already exposed. Muse lets Meta train models on user input by default, with an opt-out; Meta can still access data in the Muse virtual machine. Muse also had a zero-day flaw that allowed an agent to be taken over, now patched. The report gives no details on Dots' data controls or exposed problems.
Written by AI from the coverage · updated 1 hour ago
Coverage
Follow the reports to see the story from different sides.
- The Verge · AIAI agent makers are promising privacy — will they deliver?
OpenAI 的 Dots 与 Meta 的 Muse 都以隐私安全吸引用户，但两者的承诺、数据控制措施与已暴露的问题存在差异。Muse 默认允许 Meta 使用用户输入训练模型，用户可以退出；Meta 仍能访问其虚拟机中的数据，一处可导致智能体被接管的零日漏洞已修补。
Heat over time
Not enough continuous observations to draw a trend yet.