February 20, 2026, 4:33 pm | Read time: 3 minutes
Meta has secured a patent that may disturb many people. The company is considering allowing AI to continue managing Facebook and Instagram accounts of deceased users. This is according to a U.S. patent filed in 2023 and expected to be granted by the end of 2025, as reported by Business Insider. The company is contemplating not letting social media accounts rest after death but instead using them technically.
The idea behind this: An AI would be trained with a user’s existing data, including previous posts and direct messages. Based on this, a bot could later remain active in the name of the deceased person, giving likes and writing comments. Furthermore, the technology could enable audio and video calls using deepfakes. Deepfakes are audio or video recordings created with artificial intelligence that appear so realistic they are almost indistinguishable from real recordings.
AI to Accurately Mimic Writing Style
According to Meta, the AI would be capable of closely imitating a user’s communication style. It would analyze previous posts to derive typical phrases, grammatical nuances, and personal quirks. This could create the impression that the deceased person is still writing posts or responding to messages.
Also of interest: Meta lawsuit could bring users thousands of euros
Besides potential applications for grief management for survivors, the report also sees economic aspects. Such a feature could help maintain user numbers and interactions on the platforms, even if the number of active users declines. At the same time, the impact of similar approaches is considered controversial. According to the report, Meta CEO Mark Zuckerberg warned in 2023 about potentially harmful effects of interactions with virtual memories.
No Implementation Planned Currently
Those fearing zombie chatbots on social networks can breathe a sigh of relief for now. In response to an inquiry from Business Insider, Meta stated that there are currently no plans to turn the patent into a product. Nevertheless, the patent indicates potential commercial viability.
If implementation were to occur, numerous legal and ethical questions would need to be addressed. British professor Edina Harbinja told Business Insider that it would need to be determined who owns the data generated in the name of the deceased. Privacy concerns also play a role, even after death. Additionally, it is currently difficult to assess how authentic AI-generated posts would actually be.
Meta also emphasizes that the technology might not only be intended for deceased users. It could also be used for people who cannot or do not want to use their account for an extended period. Influencers and content creators, in particular, often lose reach during longer breaks. In such cases, an AI could produce content to maintain their presence.