- Joined
- Jan 22, 2026
- Messages
- 101
- Reaction score
- 943
It seems that now you can't even leave Facebook upside down.
Meta* has received a patent for a technology that, with the help of AI, can simulate human activity on social networks so that the account continues to "live" even after the death of the owner. The patent describes a scenario where a large language model learns from "user" data and reproduces a familiar style of behavior: likes, writes comments, responds in private messages and maintains the feeling that the person is still online. At the same time, a Meta* representative told Business Insider that the company "has no plans to advance this idea further," and recalled that obtaining a patent does not necessarily mean launching a product.
The logic of the patent itself is quite mundane: when a user disappears from the feed for a long time, the experience of his subscribers changes. And if he died, the breakup becomes "much more severe and final" because the person will not return to the platform. As a solution, Meta suggests creating a digital "snapshot" of social activity based on the history of publications, reactions, and interactions. The patent specifically states that the model can "simulate a user when he is absent from a social network, for example, if he takes a long break or if he has died." The patent was filed in 2023, and one of the key authors is CTO Meta Andrew Bosworth.
This concept has a more practical application. For bloggers and content creators who earn money on Meta* platforms, the tool could maintain engagement while a person is on hiatus or on vacation. The patent also mentions ideas about simulating audio or video calls, that is, not only text communication, but also more "live" presence formats.
But the closer the scenario gets to a postmortem imitation, the more questions there are. Edina Harbinha, a law professor at the University of Birmingham who specializes in digital rights and postmortem privacy, warns that legal, social, ethical, and even philosophical issues are mixed here. And it's not just about who consents to a "digital twin," who manages it after death, and where the boundaries of privacy lie, but also about how such systems change the very experience of loss.
Meta has thought about the "digital legacy" before. About ten years ago, Facebook* launched functions like "heir's contact" to manage the deceased's page. And in 2023, Mark Zuckerberg, in a conversation with podcaster Lex Friedman, talked about virtual avatars of the dead and how the ability to "interact or relive memories" could potentially help people in grief. The new patent continues this line, but with a much stronger bet on generative AI.
In parallel, the whole market of memory technologies (grief tech) is developing: Startups make "memory bots" and "ghost" assistants, often from the personal experience of the founders. Business Insider mentions the examples of Replicka (founded in 2015) and You, Only Virtual (founded in 2020), and also recalls that in 2021 Microsoft also patented a chatbot that could mimic a deceased person. Sociologist Joseph Davis from the University of Virginia fears that such tools can confuse the process of living with loss: one of the tasks of grief is to recognize the real loss, and the "return" of a person in the form of a convincing imitation can blur this line.
A separate layer of questions concerns the motivation of platforms. Harbinha notes that for the social network, this could mean more engagement, more content, and more data, including for training future ais. And even if Meta is now saying that it is not going to deploy the technology, the patent itself shows which way major platforms are thinking about the "immortality" of accounts and where memory ends and simulation begins.
Instagram Facebook, Threads, and Meta companies and their products have been recognized as extremist and their activities are prohibited in the Russian Federation.
Meta* has received a patent for a technology that, with the help of AI, can simulate human activity on social networks so that the account continues to "live" even after the death of the owner. The patent describes a scenario where a large language model learns from "user" data and reproduces a familiar style of behavior: likes, writes comments, responds in private messages and maintains the feeling that the person is still online. At the same time, a Meta* representative told Business Insider that the company "has no plans to advance this idea further," and recalled that obtaining a patent does not necessarily mean launching a product.
The logic of the patent itself is quite mundane: when a user disappears from the feed for a long time, the experience of his subscribers changes. And if he died, the breakup becomes "much more severe and final" because the person will not return to the platform. As a solution, Meta suggests creating a digital "snapshot" of social activity based on the history of publications, reactions, and interactions. The patent specifically states that the model can "simulate a user when he is absent from a social network, for example, if he takes a long break or if he has died." The patent was filed in 2023, and one of the key authors is CTO Meta Andrew Bosworth.
This concept has a more practical application. For bloggers and content creators who earn money on Meta* platforms, the tool could maintain engagement while a person is on hiatus or on vacation. The patent also mentions ideas about simulating audio or video calls, that is, not only text communication, but also more "live" presence formats.
But the closer the scenario gets to a postmortem imitation, the more questions there are. Edina Harbinha, a law professor at the University of Birmingham who specializes in digital rights and postmortem privacy, warns that legal, social, ethical, and even philosophical issues are mixed here. And it's not just about who consents to a "digital twin," who manages it after death, and where the boundaries of privacy lie, but also about how such systems change the very experience of loss.
Meta has thought about the "digital legacy" before. About ten years ago, Facebook* launched functions like "heir's contact" to manage the deceased's page. And in 2023, Mark Zuckerberg, in a conversation with podcaster Lex Friedman, talked about virtual avatars of the dead and how the ability to "interact or relive memories" could potentially help people in grief. The new patent continues this line, but with a much stronger bet on generative AI.
In parallel, the whole market of memory technologies (grief tech) is developing: Startups make "memory bots" and "ghost" assistants, often from the personal experience of the founders. Business Insider mentions the examples of Replicka (founded in 2015) and You, Only Virtual (founded in 2020), and also recalls that in 2021 Microsoft also patented a chatbot that could mimic a deceased person. Sociologist Joseph Davis from the University of Virginia fears that such tools can confuse the process of living with loss: one of the tasks of grief is to recognize the real loss, and the "return" of a person in the form of a convincing imitation can blur this line.
A separate layer of questions concerns the motivation of platforms. Harbinha notes that for the social network, this could mean more engagement, more content, and more data, including for training future ais. And even if Meta is now saying that it is not going to deploy the technology, the patent itself shows which way major platforms are thinking about the "immortality" of accounts and where memory ends and simulation begins.
Instagram Facebook, Threads, and Meta companies and their products have been recognized as extremist and their activities are prohibited in the Russian Federation.
