Can AI truly bring back the dead? The rise of 'Grief Tech' and its ethical dilemmas

It usually starts with something small. A song that suddenly plays on shuffle, a voice note you accidentally stumble upon, or a photograph that catches you off guard on a random Tuesday afternoon.

For a few seconds, the memory feels real and almost within reach. Now imagine if that feeling didn’t have to fade. Imagine if the song kept playing, from an urn holding their ashes. Or if their voice could respond when you spoke and advice you when you need it the most. Not from memory, but in real time. That’s the strange, slightly surreal space we’re entering. With Spotify launching “Eternal Playlist Urn”, memory has quite literally found a speaker.

It’s a Bluetooth-enabled urn that lets loved ones play a person’s favourite playlists even after they’re gone. It sounds gimmicky at first, but it also opens the door to something much bigger: The idea that maybe death doesn’t have to feel so final anymore.

Is death really the end?

And it’s not just one product. There’s an entire wave of what is now being called “grief tech”. Platforms like HereAfter AI and StoryFile allow people to create digital versions of themselves that can live on even after death.

AI tools can now recreate voices with startling accuracy, using companies like ElevenLabs or OpenAI’s voice tech. There are chatbots that simulate conversations with the deceased, and even tools that can animate old photographs or place loved ones into new images.

It’s eerie, yes—but also oddly comforting. Because grief, at its core, is about wanting just a little more time. Clinical psychologist Chetna Luthra, Marengo Asia Hospitals, Gurugram, explains it in a way that feels instantly relatable, “Grief is not about letting go completely,” she adds, “It’s about transforming the relationship.”

In that sense, these technologies can feel like a bridge. They allow people to say things they never got to say, revisit memories, or simply sit with the illusion of one’s presence for a while.

Especially in the early days of loss, when the shock is still fresh, that can feel like relief.

But here’s where it gets complicated

Because there’s a quiet difference between remembering someone and interacting with them. One lives in your head whereas the other feels like it’s happening right in front of you.

And that difference, experts say, can change how we process loss. Dr. Mahendra Mane, psychiatrist at Inamdar Hospital, Pune, says, “holding on to memories, rituals, and symbolic connections to deceased loved ones is normal and healthy.

However, it is also necessary for the griever to gradually accept the reality of the loss. If the griever is using this technology to avoid accepting reality, then it could interfere in their emotional well-being.”

In simple terms, it may comfort you, but it may also keep you stuck. Luthra adds another layer to this. She says, “The final stage of grief involves knowing one is gone and the relationship continues internally however with the AI the reality becomes complex since they are interactable where only partial acceptance comes in.”

When we remember someone, our brain knows they’re gone, even if it hurts. But when an AI version responds—talks, reacts, even surprises you—it can blur that boundary.

The brain starts to feel like the person is still accessible. And that can make it harder to fully process their absence. Over time, what begins as comfort can turn into dependence. The technology becomes the place you go to when the loneliness hits. It soothes, but it also delays. Like a painkiller that numbs the ache without really healing it.

And yet, it’s not entirely a bad thing

For some, these tools genuinely help. They offer a sense of continuity, a softer landing into grief.

Not everyone wants closure in the traditional sense. Sometimes, people just want connection, even if it’s simulated. Which is perhaps why this space is quietly turning into a business opportunity.

Startups like HereAfter AI and StoryFile are already building products around digital memory and posthumous interaction. Big tech platforms like Meta and Google have introduced features to manage accounts after death.

As our lives become increasingly digital, it’s only natural that our afterlives follow suit. There’s a clear demand! And where there’s demand, there’s a market. Manish Mohta, founder of Learning Spiral.AI, notes that AI systems today can convincingly replicate voice and conversational style using relatively small amounts of data. But he is quick to add that these are still simulations—not consciousness. “The system generates outputs that appear genuine during brief interactions, yet users will find them unsatisfactory because they lack authentic self-awareness and emotional depth, which makes them seem real but fails to capture a complete human identity.

”

Mohta adds, “Users who fail to differentiate between real interactions and simulated ones will experience deception and emotional manipulation, and they will lose trust. Organizations implement transparency measures by using specific methods, which include marking all AI systems and defining their operational limits.” Yet, for many users, the distinction may not matter. But the business side of grief brings its own set of uncomfortable questions. Like, who owns your digital self after you’re gone? Monica Lakhanpal, a legal expert and founder of VakeelSaab, says the law hasn’t quite caught up yet.

In India, while the Digital Personal Data Protection Act (DPDPA) allows individuals to nominate someone to manage their data, it doesn’t clearly define ownership of digital assets after death. In reality, control is split between platforms, families, and companies offering these AI services. “The DPDP Act is a significant step forward for data governance, but it is architecturally designed around living.

It defines a "data principal" as the individual to whom data pertains; once that individual dies, the Act's protections do not automatically transfer to the estate or family.

The data enters a void,” she says.

Consent is another grey area. Just because someone shared their data while alive doesn’t mean they agreed to be recreated after death. And yet, in many cases, that line is blurred or simply ignored. Monica notes that globally, too, regulations are patchy. She adds, “The EU's GDPR explicitly excludes deceased persons under Recital 27, leaving member states to legislate independently which most have not done.

The EU AI Act mandates transparency labelling for synthetic media but does not address posthumous simulation. California's AB 2602 (2024) requires prior written consent for AI replication of a deceased performer's voice or likeness, but no equivalent protection exists for ordinary citizens.

France remains the most progressive jurisdiction: its 2016 Digital Republic Act allows individuals to leave binding instructions for data deletion after death.”

What makes this moment feel so strange

On one hand, there’s something deeply human about wanting to hold on. To hear a familiar voice. To feel like someone is still around, even if just for a few minutes. There’s beauty in that. A kind of quiet, personal comfort that technology is now able to provide in ways we couldn’t have imagined before.

But on the other hand, grief has always been about learning to live with absence. About slowly, painfully, accepting that some conversations are over even when they are unfinished. That some doors don’t reopen. But what happens when those doors appear to open again? We’re not just changing how we remember people. We’re changing what it means to lose them. And maybe that’s the real story here. Not just about technology or innovation, but about us—our growing discomfort with endings, our desire to stretch moments just a little longer, to keep people close in whatever way we can. Because if there’s one thing this new wave of grief tech reveals, it’s this: saying goodbye has never been easy. We’re just finding new ways to avoid it.