# Babylon Health, the AI-first primary-care company, collapsed into bankruptcy in 2023 and sold its UK service for $620,000

> Babylon Health marketed an AI symptom-checker chatbot as able to diagnose as well as doctors — it claimed 82 percent on GP-exam questions against a 72 percent doctors' average — went public through a SPAC merger with Alkuri (agreement dated June 3, 2021) and began trading on the NYSE in October 2021, then filed for Chapter 7 bankruptcy in Delaware in August 2023; its UK subsidiary went into administration and substantially all its UK telemedicine assets were sold to eMed for just $620,000 (£500,000).

- Verification status: pending
- Case type: deployment
- Provider: Babylon Health (Babylon Holdings Ltd)
- Client: Babylon Health (Babylon Holdings Ltd), consumer health (named)
- Sector: consumer health / UK / cross
- Canonical URL: https://theinternetninja.com/stories/babylon-health-ai-first-primary-care-spac-collapse-chapter-7-bankruptcy-2023/
- Source: The Internet Ninja (theinternetninja.com), independent verified-proof platform

## Verification method

Every critical figure was read verbatim this session and byte-verified in sources/. Each of the three critical claims now carries a Tier-1 primary alongside the independent press: the AI-vs-doctors accuracy claim is documented and rebutted in the peer-reviewed Lancet correspondence (Fraser, Coiera, Wong, 2018); the SPAC combination's equity values and dates come from Babylon's own SEC 424B3 and 6-K Ex-99.1 filings; and the Chapter 7 collapse is on the U.S. Bankruptcy Court (D. Delaware) record via CourtListener/RECAP (cases 23-11127 and 23-11128, both filed 2023-08-09). The 424B3 reconciles the SPAC valuations the press reported separately ($4.2 billion post-closing equity value and $3.515 billion current equity value are two figures from one filing, not a conflict). The exact 82%/72% split (Babylon's own claim, relayed by CNBC) and the $620,000 (£500,000) eMed sale price (Forbes/hlth.com, TechCrunch) remain Tier-2.

## Full case file

## The problem
Babylon Health built its pitch on an AI symptom-checker chatbot marketed as a substitute for seeing a doctor, and its central claim was quantitative: at a June 2018 launch event the company said its AI could "diagnose patients better than humans can," a claim doctors called "dubious" ([source](https://www.cnbc.com/2018/06/28/babylon-claims-its-ai-can-diagnose-patients-better-than-doctors.html)). The comparison Babylon offered was that "the average pass mark for real-life doctors over the past five years was 72 percent, according to Bablyon. Its AI, however, scored 82 percent" ([source](https://www.cnbc.com/2018/06/28/babylon-claims-its-ai-can-diagnose-patients-better-than-doctors.html)). The claim was framed as the AI diagnosing "common diseases as well as human physicians" ([source](https://undark.org/2019/12/09/babylon-health-artificial-intelligence-medical-advice/)), and the independent scientific record was unconvinced. In a correspondence in The Lancet, researchers Fraser, Coiera and Wong recorded that "the health technology company Babylon recently claimed that their Babylon Diagnostic and Triage System outperformed the average human doctor on a subset of the Royal College of General Practitioners exam" ([source](https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)32819-8/fulltext)), and concluded that "Babylon's study does not offer convincing evidence that its Babylon Diagnostic and Triage System can perform better than doctors in any realistic situation, and there is a possibility that it might perform significantly worse" ([source](https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)32819-8/fulltext)).

## What was built
The product was an AI-first digital primary-care operation: a triage-and-diagnosis chatbot sold direct to consumers and offered through the NHS as the "GP at Hand" service. The safety of that chatbot was publicly contested before any financial trouble. Consultant oncologist Dr David Watkins, of "The Royal Marsden NHS Foundation Trust," spent months documenting diagnostic failures in the tool, and rather than engage the concerns Babylon issued a press release publicly attacking him ([source](https://techcrunch.com/2020/02/25/first-do-no-harm/)). The weakest load-bearing source in this record is that clinician-safety strand: it rests on a single newsroom (TechCrunch) reporting one doctor's testing, so it is treated here as supporting context, not as a critical figure.

## The outcome
Babylon took the AI story to the public markets through a SPAC merger with Alkuri Global Acquisition Corp. Its own SEC prospectus states the deal "implies a $4.2 billion post-closing equity value and a current equity value of Babylon at $3.515 billion" ([source](https://www.sec.gov/Archives/edgar/data/1866390/000119312521288057/d139720d424b3.htm)), under a merger agreement "dated as of June 3, 2021" ([source](https://www.sec.gov/Archives/edgar/data/1866390/000119312521288057/d139720d424b3.htm)). That first-party filing reconciles two figures reputable press reported separately, rather than contradicting them: Forbes's "$4.2 billion valuation" ([source](https://hlth.com/insights/news/bankrupt-digital-health-company-babylon-sells-u-k-assets-for-just-620-000-2023-09-20)) is the post-closing equity value and Healthcare Dive's "the SPAC deal valued Babylon at $3.5 billion" ([source](https://www.healthcaredive.com/news/Babylon-Chapter-7-bankruptcy/691218/)) is the $3.515 billion current equity value from the same document. The listing itself came later than the "June 2021" press shorthand implies: in a press release dated "21 October 2021" Babylon announced that it had "completed its business combination with Alkuri Global Acquisition Corp" and that its Class A ordinary shares and warrants "will begin trading on the New York Stock Exchange ("NYSE") on October 22, 2021 under the new ticker symbols "BBLN"" ([source](https://www.sec.gov/Archives/edgar/data/1866390/000119312521304150/d231170dex991.htm)). The 21 October date is the announcement dateline; trading began the following day, 22 October 2021. By the time it collapsed, TechCrunch described "the London telehealth startup once valued at nearly $2 billion" ([source](https://techcrunch.com/2023/08/31/the-fall-of-babylon-failed-tele-health-startup-once-valued-at-nearly-2b-goes-bankrupt-and-sold-for-parts/)). The collapse was fast, and it is on the court record: in August 2023 Babylon filed for Chapter 7 in the U.S. Bankruptcy Court for the District of Delaware for two subsidiaries, Babylon Inc. (case 23-11127) and Babylon Healthcare Inc. (case 23-11128), both docketed as Chapter 7 filings dated 2023-08-09 ([source](https://www.courtlistener.com/docket/67680026/babylon-inc/)) ([source](https://www.courtlistener.com/docket/67680024/babylon-healthcare-inc/)). Healthcare Dive reported the same event as "Babylon filed for bankruptcy in Delaware last week for two of its subsidiaries, according to court filings" ([source](https://www.healthcaredive.com/news/Babylon-Chapter-7-bankruptcy/691218/)), and its UK subsidiary "went into administration" ([source](https://techcrunch.com/2023/08/31/the-fall-of-babylon-failed-tele-health-startup-once-valued-at-nearly-2b-goes-bankrupt-and-sold-for-parts/)). What remained of the UK business was worth almost nothing: its core UK telemedicine service was "sold for just $620,000 (£500,000) to U.S. digital health business eMed." ([source](https://hlth.com/insights/news/bankrupt-digital-health-company-babylon-sells-u-k-assets-for-just-620-000-2023-09-20)). The finances underneath were deeply negative — Babylon reported a "$63.2 million net loss" in Q1 2023 ([source](https://www.healthcaredive.com/news/Babylon-Chapter-7-bankruptcy/691218/)), its Delaware filings "list liabilities between $100 million and $500 million" ([source](https://www.healthcaredive.com/news/Babylon-Chapter-7-bankruptcy/691218/)), and by the end it "owes AlbaCore a combined $380.5 million in loans" ([source](https://hlth.com/insights/news/bankrupt-digital-health-company-babylon-sells-u-k-assets-for-just-620-000-2023-09-20)).

## Why this belongs in the record
This is an AI-hype walk-back, not a positive vendor outcome: the headline figures are a bankruptcy and a fire-sale, and there are no outcomes to badge. Its value in the corpus is as a documented arc from a specific, quantified AI claim — an 82-percent-versus-72-percent diagnostic boast that independent clinicians and The Lancet said the evidence did not support — to a company that raised billions on that story and then liquidated, selling its flagship AI-first service for $620,000. It is the medical counterpart to the algorithmic bets that were unwound at scale once reality diverged from the model.

## How this was verified
- **Method:** Every quoted figure and holding was read verbatim from sources captured this session to web.archive.org and copied into `sources/`. No number is paraphrased, rounded, or recalled; where the SPAC valuation conflicts across outlets, all three figures are shown, not merged.
- **Tier-1 primaries, with independent press alongside:** each critical claim carries a first-party or primary-record source fetched and archived this session. The AI-accuracy claim rests on the peer-reviewed Lancet correspondence (Tier 1), which records Babylon's RCGP-exam outperformance claim verbatim and rebuts it, corroborated by CNBC (2018) and Undark (2019). The SPAC combination rests on Babylon's own SEC 424B3 and 6-K Ex-99.1 filings (Tier 1), which pin the equity values, the June 3, 2021 merger-agreement date and the 22 October 2021 start of NYSE trading (announced 21 October 2021), corroborated by Forbes and Healthcare Dive. The Chapter 7 collapse rests on the U.S. Bankruptcy Court (D. Delaware) record via CourtListener/RECAP (Tier 1), corroborated by Healthcare Dive and TechCrunch. The exact 82%/72% figures (Babylon's own claim via CNBC) and the $620,000 eMed sale price (Forbes/hlth.com, TechCrunch) remain Tier-2, and are named as such.
- **Conflation guards:** the $63.2M Q1-2023 net loss (a quarterly figure) is never merged with the $380.5M owed to AlbaCore (total loans) or the $100M–$500M liabilities range (the Delaware filings' stated band); the three SPAC valuations are held as a disclosed conflict, not averaged.
- **Verified on:** 2026-09-03 (maker research pass; handed to the checker at `checking`).

## Related case files
- [Epic's sepsis-prediction model missed most sepsis cases in external validation](/stories/epic-sepsis-model-external-validation-missed-67pct-sepsis-auc-063-jama-michigan-2021/) — the same gap between a healthcare-AI vendor's accuracy claims and its independently measured real-world performance.
- [Zillow shut its algorithmic home-buying business after a $304M write-down](/stories/zillow-offers-ibuying-algorithm-shutdown-304m-writedown-25pct-layoffs-2021/) — another company that raised the AI story to scale and unwound the bet at a heavy loss when the model underperformed.
- [Humane's AI Pin discontinued; HP buys the assets and bricks the devices](/stories/humane-ai-pin-discontinued-hp-116m-acquisition-devices-bricked-2025/) — a flagship AI product retired for a fraction of its raised capital, the closest sibling for "the AI bet sold for parts."
- [Klarna reversed its AI-first customer service and began hiring humans again](/stories/klarna-ai-customer-service-walkback-rehiring-humans-2025/) — an AI-first strategy walked back once quality problems surfaced, the operational counterpart to Babylon's collapse.

## Sources
1. **[Tier 1]** The Lancet · "Safety of patient-facing digital symptom checkers" (Hamish Fraser, Enrico Coiera, David Wong) · November 6, 2018 · https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)32819-8/fulltext
2. **[Tier 1]** U.S. SEC · Babylon Holdings Ltd, Form 424B3 (Alkuri Global Acquisition Corp business-combination prospectus) · September 30, 2021 · https://www.sec.gov/Archives/edgar/data/1866390/000119312521288057/d139720d424b3.htm
3. **[Tier 1]** U.S. SEC · Babylon Holdings Ltd, Form 6-K Exhibit 99.1 (NYSE-listing press release, ticker BBLN) · October 21, 2021 · https://www.sec.gov/Archives/edgar/data/1866390/000119312521304150/d231170dex991.htm
4. **[Tier 1]** U.S. Bankruptcy Court, District of Delaware (via CourtListener/RECAP) · Babylon Inc., case 23-11127, Chapter 7 · filed August 9, 2023 · https://www.courtlistener.com/docket/67680026/babylon-inc/
5. **[Tier 1]** U.S. Bankruptcy Court, District of Delaware (via CourtListener/RECAP) · Babylon Healthcare Inc., case 23-11128, Chapter 7 · filed August 9, 2023 · https://www.courtlistener.com/docket/67680024/babylon-healthcare-inc/
6. **[Tier 2]** CNBC · "Babylon claims its AI can diagnose patients better than doctors, doctors call that 'dubious'" · June 28, 2018 · https://www.cnbc.com/2018/06/28/babylon-claims-its-ai-can-diagnose-patients-better-than-doctors.html
7. **[Tier 2]** Undark · "Medical Advice From a Bot: The Unproven Promise of Babylon Health" · December 9, 2019 · https://undark.org/2019/12/09/babylon-health-artificial-intelligence-medical-advice/
8. **[Tier 2]** TechCrunch · "AI chatbot maker Babylon Health attacks clinician in PR stunt after he goes public with safety concerns" · February 25, 2020 · https://techcrunch.com/2020/02/25/first-do-no-harm/
9. **[Tier 2]** Healthcare Dive · "Babylon sells UK business as digital health company winds down operations" · August 16, 2023 · https://www.healthcaredive.com/news/Babylon-Chapter-7-bankruptcy/691218/
10. **[Tier 2]** Forbes (syndicated on hlth.com) · "Bankrupt Digital Health Company Babylon Sells U.K. Assets For Just $620,000" · September 20, 2023 · https://hlth.com/insights/news/bankrupt-digital-health-company-babylon-sells-u-k-assets-for-just-620-000-2023-09-20
11. **[Tier 2]** TechCrunch · "The fall of Babylon: Failed telehealth startup once valued at nearly $2B goes bankrupt and sold for parts" · August 31, 2023 · https://techcrunch.com/2023/08/31/the-fall-of-babylon-failed-tele-health-startup-once-valued-at-nearly-2b-goes-bankrupt-and-sold-for-parts/