• Meta secured a patent for AI that could replicate a user’s social media behavior after death, including posts, messages and simulated calls
• Privacy experts and grief researchers warn the technology could exploit mourning families while boosting platform engagement metrics
• The patent joins a growing “grief tech” industry that has drawn scrutiny from Cambridge ethicists and remains largely unregulated

MENLO PARK, CA (TDR) — Meta has been granted a patent describing an artificial intelligence system capable of simulating a user’s social media activity after extended absence or death, igniting a fierce debate among technologists, ethicists and grief counselors over where the line falls between digital legacy and commercial exploitation.

The patent, first filed in November 2023 and granted in late December 2025, outlines a large language model trained on a user’s historical online activity — posts, comments, likes, messages and voice data — to replicate their behavior across Meta’s platforms. The filing lists Andrew Bosworth, Meta’s chief technology officer, as the primary inventor.

What the Meta AI Patent Actually Describes

The patent goes well beyond simple account memorialization. According to the filing, the system could generate posts, respond to direct messages, comment on content and interact with other users in patterns that mirror the original account holder’s established behavior. The document also references advanced capabilities including simulated audio and video calls using the reconstructed digital persona.

“The language model may be used for simulating the user when the user is absent from the social networking system, for example, when the user takes a long break or if the user is deceased.”

Freedom-Loving Beachwear by Red Beach Nation - Save 10% With Code RVM10

The patent argues that a user’s sudden disappearance from a platform affects the broader community of connected accounts.

“The impact on the users is much more severe and permanent if that user is deceased and can never return to the social networking platform.”

A Meta spokesperson emphasized the filing does not represent a product announcement.

“We have no plans to move forward with this example.”

CLICK HERE TO READ MORE FROM THE THE DUPREE REPORT

The spokesperson added that companies routinely file patents to protect or disclose concepts that may never be developed into commercial products.

The Business Incentive Critics Can’t Ignore

While Meta frames the patent around user experience and community continuity, several researchers have pointed to a more straightforward calculation: engagement. Edina Harbinja, a professor at the University of Birmingham Law School specializing in digital rights and post-mortem privacy, identified the financial logic directly.

“It’s more engagement, more content, more data — more data for the current and the future AI. I can see the business incentive for that.”

Facebook has accumulated an estimated 30 million or more accounts belonging to deceased users over its two decades of operation — a growing reservoir of inactive profiles that generate no ad revenue and no engagement metrics. Reactivating even a fraction of those accounts through AI simulation could represent significant commercial value.

Harbinja cautioned that companies must balance innovation against the digital rights of individuals after death and questioned who ultimately controls a person’s digital representation once they’re no longer able to consent.

“It raises social, ethical, and philosophical questions about how people are represented after death and who controls that representation.”

Grief Researchers Sound the Alarm

The psychological stakes have drawn attention from academics who study how technology intersects with mourning. Joseph Davis, a sociology professor at the University of Virginia, expressed direct concern about what AI simulations could do to the grieving process.

“One of the tasks of grief is to face the actual loss. Let the dead be dead.”

Davis argued that blurring the boundary between presence and absence could interfere with healthy mourning rather than support it.

Researchers at Cambridge University’s Leverhulme Centre for the Future of Intelligence published a peer-reviewed study in Philosophy and Technology that outlined three scenarios where AI-powered “deadbots” could cause harm — from surreptitiously advertising products in a deceased grandmother’s voice to distressing children by insisting a dead parent remains present.

“This area of AI is an ethical minefield. It’s important to prioritize the dignity of the deceased, and ensure that this isn’t encroached on by financial motives of digital afterlife services.”

Katarzyna Nowaczyk-Basińska, a Cambridge research fellow and co-author of the study, warned that the technology is already outpacing the ethical frameworks needed to govern it.

“We need to start thinking now about how we mitigate the social and psychological risks of digital immortality, because the technology is already here.”

The Cambridge researchers recommended age restrictions, meaningful transparency measures and “digital funeral” protocols that would allow families to retire AI simulations of their loved ones in dignified ways.

Meta AI Patent Isn’t Without Precedent

Meta is far from the first technology company to explore this territory. Microsoft secured a similar patent in 2021 for a chatbot capable of simulating deceased individuals, fictional characters or celebrities using social media data, images and voice recordings. Tim O’Brien, then Microsoft’s general manager of AI programs, acknowledged publicly that the concept was “disturbing” and confirmed Microsoft had no plans to pursue it.

Several startups have already moved into the space. Companies like Eternos, HereAfter AI and You, Only Virtual offer services that create interactive digital avatars of deceased loved ones. Project December, which initially harnessed GPT models before developing its own systems, gained attention when a user trained it to simulate conversations with his deceased fiancée.

The critical difference between those services and Meta’s patent, according to experts, is consent. Existing grief tech startups typically require active participation from surviving family members. Meta’s filing describes a system that could theoretically activate automatically based on account inactivity — raising questions about whether users or their families would meaningfully opt into having an AI clone continue their digital presence.

Zuckerberg Previously Floated the Concept

The patent aligns with comments Mark Zuckerberg made in a 2023 interview with podcaster Lex Fridman, where the Meta CEO suggested AI could eventually help people process grief through digital interactions with representations of the deceased.

“If someone has lost a loved one and is grieving, there may be ways in which being able to interact or relive certain memories could be helpful. But then there’s also probably an extent to which it could become unhealthy.”

Zuckerberg acknowledged Meta already had significant experience handling death and digital identity through its existing memorialization features, which allow users to designate a “legacy contact” who can manage limited aspects of their profile after death.

“We have, you know, a fair amount of experience with how to handle death and identity and people’s digital content through social media already, unfortunately.”

The Regulatory Vacuum

The patent arrives during a period of rapidly evolving AI regulation — but one that has yet to address grief technology specifically. The European Union’s AI Act, which entered into force in August 2024, classifies certain AI applications as high-risk and imposes transparency requirements on systems that interact with humans. Key provisions around high-risk AI systems are scheduled to take full effect by August 2026.

Under the EU framework, AI systems designed to simulate human interaction would likely face disclosure obligations requiring platforms to inform users when they’re communicating with a machine rather than a person. Whether a deceased user’s AI-generated posts would trigger those requirements remains an open legal question.

In the United States, no comparable federal framework governs AI-powered simulations of the deceased. Dr. Tomasz Hollanek, co-author of the Cambridge study, stressed that the gap between what is technically possible and what is legally governed continues to widen.

“People might develop strong emotional bonds with such simulations, which will make them particularly vulnerable to manipulation. Methods and even rituals for retiring deadbots in a dignified way should be considered.”

The lack of clear regulation means companies filing patents in this space face few legal constraints beyond existing data privacy laws — which vary dramatically by jurisdiction and typically offer limited protections for the digital identities of the deceased.

As AI systems grow powerful enough to replicate individual behavior patterns with increasing fidelity, who should decide whether a person’s digital presence outlives them — the platform that holds their data, the family that mourns them, or the individual who never anticipated the question?

Sources

This report was compiled using information from Business Insider’s original reporting on the patent filing, analysis by Fast Company, TechSpot, and Futurism, coverage by Audacy, Cybernews, Mediaweek, and GreekReporter, the Cambridge University Leverhulme Centre’s research on deadbot safeguards, the peer-reviewed study published in Philosophy and Technology, EU AI Act implementation guidance, and reporting by The Washington Post on Microsoft’s earlier patent.

Freedom-Loving Beachwear by Red Beach Nation - Save 10% With Code RVM10