Dead man talking: pushing the limits of curated feeds

Meta has patented technology that allows artificial intelligence to pose as you after you are dead. Not preserve your words or protect your legacy, but replace you. The patent could empower Meta to use your digital remains to bait the emotions of your living loved ones.

This is not technological progress. It is a corporation stepping into the space between the living and the dead and deciding that even death should not disrupt engagement metrics. It is not innovation. It is corporate overreach into the one boundary that should remain inviolable.

The patent describes training large language models on a user’s messages, tone, photographs, and voice to simulate their presence indefinitely. Similar experiments are under way elsewhere too. Amazon has demonstrated Alexa reading in a deceased grandmother’s voice, and Microsoft once patented a chatbot designed to recreate the dead from their digital traces. Its own leadership later called the idea “disturbing”.

The world is witnessing the next phase of social media’s expansion, not into new geographies or devices, but into death itself.

When Facebook crossed one billion users in 2012, the milestone was publicly celebrated as a triumph. But in her book Careless People, former insider Sarah Wynn-Williams wrote that behind the scenes, the mood among senior leadership was far less jubilant. Growth was slowing, the stock price was wobbling, and the pressure to demonstrate dramatic expansion was mounting.

The company then focused its energy into penetrating markets such as China that were hostile to social media platforms. Now, 14 years later, it’s reaching for the afterlife.

The system is so addicted to continuity and monetisation that it refuses to let even death interrupt the feed.

For more than a decade, critics have warned social media distorts reality. In his book The Chaos Machine, writer Max Fisher argued the early conventional wisdom that these platforms simply amplify sensationalism and outrage, vastly understated their power. The technology reshapes how we think and behave. Algorithms reward moral fury and compress complexity into viral fragments.

In 2017, Facebook’s founding president Sean Parker admitted the goal was to consume as much of users’ conscious time and attention as possible. This patent, filed six years later, extends that logic beyond life itself.

There’s no debating the harms associated with the current state of social media. The introduction of “like” buttons and reaction emojis turned human expression into measurable performance. Approval became quantifiable. Identity became curated. We have learned to perform for the algorithm. The recently approved patent makes way for the same architecture to govern grief.

Static bots

Who owns a person’s voice after death? What happens when an AI replica expresses views the real individual might have outgrown if alive? Can a grieving spouse distinguish between comfort and manipulation when the interaction is engineered by engagement-maximising systems?

An AI clone trained on a person’s past posts will not grow. It will not evolve or change its mind. It will not reconcile. People don’t stop growing, but the bot will forever be static.

Grief is difficult because it forces acceptance. It requires the living to confront finality. Digital resurrection interrupts that process. It invites the bereaved to interact with a simulation that speaks in familiar cadences but lacks agency. It normalises a delusion, that death can be negotiated and presence can be manufactured.

Meta’s public principles speak of building community, protecting safety and respecting human dignity. Its Responsible AI statements emphasise transparency, accountability and preventing harm. Its Community Standards warn against impersonation and misrepresentation. How does posthumous simulation sit within this framework?

Seeking attention

Social media companies are not in the business of spiritual care. They are in the business of attention. Every additional interaction, even one with a dead user’s AI avatar, is sellable data.

There is also a broader civic cost. We already inhabit an environment saturated with synthetic media, including deepfakes, voice clones and AI-generated personas that some have even sought to date and marry. Objectivity has been eroded by misinformation and by the instability of what is real.

Digital immortality does not exist in isolation. It grows from the same incentive structure that amplifies outrage and rewards extremity. Fisher wrote how even social media insiders too began to recognise that the product itself, not merely bad actors using it, was driving polarisation.

There is a difference between assistive technology and existential substitution. Meta’s Aria Gen 2 glasses help people with memory loss navigate daily life. They help people remember tasks. That is augmentation in service of the living. Simulating a deceased person’s consciousness doesn’t aid, only exploits.

Digital immortality prioritises commercial continuity over closure. A society that cannot allow its dead to rest risks losing its grip on reality. Grief should not become monetisable. Technology can extend life but death should not become another product category.

Published – March 04, 2026 12:20 am IST