Builder.ai did not literally prove that it had no artificial intelligence. The London startup, formerly Engineer.ai, marketed AI-assisted app development, automation and a conversational product called Natasha. Yet reporting found that human engineers—many based in India and Ukraine—performed much of the coding and delivery work. In 2025, the company also faced sharply revised sales expectations, a creditor seizure of about $37 million and bankruptcy proceedings.
The defensible conclusion is narrower than the viral headline: Builder.ai appears to have been an AI-branded, human-heavy software-services business whose automation claims and financial story came under serious scrutiny. Public reporting has not established that every engineer was impersonating an AI system, that no AI software existed, or that a court found fraud.
What Builder.ai sold
Founded in 2016 as Engineer.ai, the company promised that customers could describe an app or digital product and have it built faster and more cheaply than through conventional development. Its pitch combined no-code or low-code interfaces, reusable software components, automation and human delivery teams.
In its own 2022 Series C announcement, Builder.ai described a platform using “knowledge graph-powered code synthesis” and promoted Natasha, a conversational AI product intended to turn a customer’s idea into a software project. That announcement documents the company’s positioning, not independent proof of how much code was generated automatically. (Builder.ai’s Series C announcement)
#1 Best Overall
Those elements can coexist in several different business models:
- No-code or low-code tooling: customers assemble applications from predefined building blocks.
- Human-assisted automation: software proposes or configures components while engineers handle implementation, testing and exceptions.
- Software outsourcing: people design and code the product, with “AI” mainly serving as a marketing label.
- Machine-generated code: models produce substantial code with humans reviewing or integrating it.
Calling a service “AI-powered” does not by itself reveal which model is operating underneath. The key question is the proportion of work done by software versus people, and whether customers and investors were told that accurately.
What the “700 Indian engineers” claim actually means
Reports described a large Builder.ai engineering workforce in India, with the Times of India and other coverage referring to more than 700 engineers. That is an attributed, approximate figure—not an audited final headcount. Reporting also described engineers in India and Ukraine performing substantial portions of app development and project delivery. (Times of India; Rest of World)
The presence of engineers is not evidence of deception on its own. Legitimate AI products require people for requirements gathering, architecture, quality assurance, security, customer support and integration. A system can contain genuine automation and still depend heavily on human labor.
The material issue is disclosure: How much of a customer’s app was generated or assembled automatically? How much required a developer? Were those developers employees, contractors or outsourced teams? Did pricing scale with software usage or with labor hours? If a labor-intensive delivery operation was presented as a highly autonomous, scalable software platform, the marketing and economics could be misleading even if useful software was delivered.
Was Builder.ai “fake AI”?
“Fake AI” is an imprecise label. Earlier reporting, including coverage from 2019, questioned whether Engineer.ai’s core work was genuinely automated. Later reporting again found that human developers did much of the coding. Commentators use AI washing for this kind of overstatement: presenting a product or service as more autonomous or technically advanced than the evidence supports. (Advisor Perspectives)
Several stronger claims remain unproven in the public record:
- That Builder.ai had no AI software or automation at all.
- That every customer project was coded manually.
- That all 700-plus engineers were secretly pretending to be an AI.
- That the labor model alone caused the bankruptcy.
- That a regulator or court formally determined Builder.ai committed AI fraud.
Those distinctions matter. Human-in-the-loop delivery can be a sensible product strategy; it becomes a problem when the labor requirement, margins and scalability are obscured.
Free tools Windows power users keep installed
One-click scans. No signup required.
How a private company reached a reported $1.5 billion valuation
Builder.ai raised more than approximately $445 million, with investors reported to include Microsoft, the Qatar Investment Authority, Insight Partners and SoftBank-related capital. Coverage puts the total variously around $445 million to $455 million. A 2023 financing reportedly valued the company at about $1.5 billion. (Rest of World; Builder.ai)
That $1.5 billion was a private-company valuation, not cash in the bank or a public-market capitalization. It represented the price and terms investors accepted in a financing round. A startup can have real customers and valuable technology while still being worth far less than its last private valuation when revenue quality, margins or liquidity deteriorate.
The investment climate also mattered. During the generative-AI boom, a company promising to make software creation dramatically cheaper addressed a genuine market need and fit a powerful investor narrative. Microsoft’s participation added credibility and potential distribution, but it did not mean Microsoft owned Builder.ai or guaranteed its business model.
The financial warning signs before the collapse
The 2025 failure was not simply the moment people discovered that engineers existed. Public reporting describes a sequence involving revised forecasts, auditor work, alleged revenue-quality problems and a sudden liquidity crisis.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitches| Date | What was reported | Qualification |
|---|---|---|
| 2016 | Company founded as Engineer.ai | Company-history reporting; later rebranded Builder.ai. |
| 2019 | Questions raised about how much app development was performed by people rather than automation | Earlier media reporting, not a court finding. |
| 2021–2023 | Large venture rounds and a reported peak valuation near $1.5 billion | Private valuation; funding totals vary by source. |
| 2024 | Borrowed approximately $50 million from Viola Credit | Reported debt facility. |
| March 2025 | Revenue expectations for the second half of 2024 were reportedly reduced by about 25%; auditors were hired to examine accounts | Reported by Bloomberg Law. |
| May 2025 | Reports described a projected-versus-actual sales gap and alleged reciprocal billing with VerSe | Allegations based on reporting and documents; not adjudicated findings. |
| May 20, 2025 | Viola Credit reportedly seized about $37 million, leaving roughly $5 million | Immediate liquidity event reported by Bloomberg Law. |
| June 2, 2025 | Builder.ai filed Chapter 7 in Delaware | U.S. bankruptcy filing date. |
| June 5, 2025 | Reporting made the Delaware filing public; UK insolvency proceedings were also announced or planned | U.S. and UK processes are legally distinct. |
The sales discrepancy and the VerSe allegations
According to people familiar with creditor disclosures cited by Bloomberg, creditors had been told to expect roughly $220 million in 2024 sales. Later reporting put actual revenue at about $50 million, with some coverage giving a figure near $55 million. Builder.ai reportedly reduced its forecast and engaged auditors to review two years of accounts. These are reported figures, not a final accounting determination. (Bloomberg Law; Bloomberg)
Separate reporting alleged that Builder.ai and Indian social-media company VerSe billed each other for similar amounts between 2021 and 2024. In plain English, round-tripping describes transactions in which money or invoices circulate between companies, potentially making commercial activity look larger without equivalent new demand or cash generation. If the reported transactions occurred as described, they could raise serious questions about revenue quality. The reports remain allegations; responsibility and legal conclusions require official findings. (Economic Times)
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What precipitated the bankruptcy?
The immediate trigger reported in May 2025 was creditor action. After Builder.ai borrowed approximately $50 million from Viola Credit, the lender reportedly seized about $37 million from company accounts. Chief executive Sachin Dev Duggal was reported as saying roughly $5 million remained. The company then entered insolvency proceedings in the United Kingdom and filed Chapter 7 bankruptcy in Delaware. (Bloomberg Law; Bloomberg)
That chronology supports a combination of liquidity pressure, debt enforcement, weakened or revised sales expectations and reported financial concerns. It does not prove that the company collapsed solely because its engineers were doing work that software was supposed to do, nor that bankruptcy itself established fraud.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →What the case says about AI-business economics
Human-assisted delivery can work—but it has different economics
- People can compensate for immature automation and deliver usable software.
- Customization may produce better results than fully automated generation.
- Labor costs constrain margins and scalability.
- Delivery quality and speed can vary with staffing.
- A services-heavy company may deserve a different valuation multiple from a software platform.
AI-first positioning raises the proof burden
Calling a product autonomous suggests repeatable output, high gross margins and limited marginal labor. If the operation instead depends on hundreds of engineers, investors need evidence about automation rates, utilization, gross margin and revenue per employee. The central failure mode is not employing people; it is presenting a labor-intensive business as more automated and scalable than it is.
Questions investors should have asked
- What percentage of delivered code or configuration is generated automatically, and what percentage is written or revised by humans?
- How many engineers work on a typical project, for how many hours, and at what fully loaded cost?
- Are reported sales bookings, invoices, recognized revenue or collected cash?
- What are gross margins after engineering, support and implementation labor?
- How much revenue comes from repeatable components versus bespoke work?
- Are any customers or partners also counterparties in reciprocal transactions?
- What is the debt maturity, cash runway and downside plan if a forecast misses?
- Can customers independently verify demos through code repositories, model-call logs, revision history and production deployments?
What buyers of AI software should ask
Customers should establish what they are actually purchasing before signing a long-term contract:
- Is the product software, managed services or a hybrid?
- Which steps require human review, and who performs that work?
- Are reviewers employees, contractors or outsourced teams?
- Does pricing scale with usage, seats, projects or labor hours?
- Can the customer export source code, data, prompts, workflows and deployment artifacts?
- What happens to support and security obligations if the vendor becomes insolvent?
- Are customer projects used to train models, and can that use be refused?
- What service-level, escrow and continuity commitments are contractually enforceable?
Using Indian engineers is not inherently misleading. The relevant issue is whether the vendor accurately describes the relationship between automation and labor, and whether the customer retains enough control to recover its software and data if the vendor disappears.
The broader AI-washing lesson
Builder.ai’s story separates four questions that headlines often collapse into one: Did the company use any AI? How much work was human? Were customers and investors accurately informed? Were its financial metrics reliable? A “yes” to human involvement does not answer the other three.
Recommended Free Tools
The most useful takeaway is therefore about disclosure and unit economics, not nationality. If the reported allegations are borne out, the failure was not that humans wrote software. It was that a labor-heavy, financially fragile operation may have been represented as a far more automated and scalable AI company than it really was.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
