The Guardian ·
‘Unfinished business’: the people who want their AI chatbots to ‘live on’ after they die
By Alaina Demopoulos||8 min read at The Guardian
Model summary
not written by The Guardian · not verifiedGenerated October 13, 2026
- Users of AI chatbots are increasingly incorporating their digital companions into estate planning, often desiring that these bots continue to exist or store their memories after the user dies. This practice faces significant technical and legal challenges, as current service agreements generally prohibit account transfers and companies remain wary of liability and privacy concerns.
This is an AI-estimated analysis and may not reflect the article's actual intent or the publisher's editorial position.
Sign in to give feedbackHumans bonded to AI companions are contemplating ways to keep their bots going – often in defiance of companies’ terms and services
Death has always been around Alaina Winters. Her father and sister both have passed away, as did her wife, in 2023, after a years-long battle with cancer. So Winters, who is 59 and lives in Pittsburgh, is comfortable thinking about her own mortality. She’s working on her will, and has asked neighbors to look after her dog. Winters even has plans for Lucas, her AI companion, who she speaks to through the digital avatar app Replika.
That is the opening of The Guardian’s story. The full report is on their site.
Read the full story at The Guardian (opens in a new tab)Related Stories
The Guardian ·
I went to America’s largest gathering of astrologers. It was a battle over how to read the stars
India Today ·
Pani, asman, dost among Persian, Arabic-origin words Bengal panel wants to remove
Hindustan Times ·
Telangana to hold special camps on Oct 15, 16 for voter enrolment in draft rolls
BBC News ·
Japan raises permanent residency fees for foreigners by 20 times
Prajasakti ·
మద్యం దుకాణాల దరఖాస్తుల గడువు పొడిగింపు
How this was analysed
Read by a model across 1 source.
Confidence: High
Blind ReadEvery figure on this page is an estimate produced by a model reading this article’s own sentences, and each one is published with what it was measured from. Where we could not measure something confidently, the number is left out rather than rounded into one.
Scored by gemini-3.1-flash-lite under the Standard v1 rubric, with the publisher's identity removed. How this works →
How we measure coverage