And it worked. 

The company’s “Natasha” agent, a front-end assistant, parsed client inputs and matched them to code templates. Claims that the service consisted of ‘82% autonomous app building’ however proved to be wildly exaggerated: Less Silicon Valley sorcery than old fashioned cheap labour, it turned out. humans—mostly based in India—were doing most of the work. Here, A.I., it seems, stood for ‘Actual Indians.’

The Wall Street Journal actually published a 2019 exposé confirming this, yet, a bit perplexingly, investors simply overlooked it. 

It found that the purported ‘AI’ amounted to rudimentary decision trees (a decades-old technology), rather than natural-language processing and the like. Pricing and scheduling functions were administered by traditional software systems, rather than by any advanced algorithms, and the bulk of development work was executed manually by a large, woefully underpaid workforce.

Employees described the remuneration as ‘terrible,’ and insiders characterised Builder.ai as ‘all humans, no intelligence’—so-called AI-washing at its best. 

It would appear then that Builder.ai had indeed made one important discovery, which may just underpin much of the AI startup market: investment in computational infrastructure, data collection and annotation, algorithm design, and the painstaking work of iterative training and model optimisation is, in fact, replaceable with a few hundred low-wage coders in some poverty-ravaged metropolis—as long as the vending machines furnishing a steady supply of caffeine and sugar remain well-stocked, at least. 

But things would only get sketchier, as a 2021 round of investment resulted in $65 million being raised, but no lead investor taking credit. Builder.AI would also inflate sales, reporting $220 million instead of its actual $55 million in 2024. At the same time, Duggal himself was becoming embroiled in an investigation in India concerning alleged money laundering, (though Builder.ai’s legal counsel maintained that he had only been summoned as a witness). 

Between 2021 and 2024, the company allegedly engaged in ‘round-tripping,’ i.e. reciprocal invoicing for services that were never rendered, with Indian social media firm VerSe Innovation—an accounting manoeuvre that artificially boosted revenue to mislead investors and creditors.

Viola Credit, a creditor that had provided $50 million in financing, froze $37 million of the company’s accounts after uncovering exaggerated financial forecasts and contract breaches. This left Builder.ai with a meagre $5 million, which, due to governmental restrictions on capital outflows, could not be allocated to payroll.

By February 2025, Duggal had resigned as chief executive, though he remained on the board under the fitting title of ‘Chief Wizard.’ His successor, Manpreet Ratia—a former Amazon and Flipkart executive who also served as managing partner at Jungle Ventures—was appointed to rectify the situation. 

To no avail, however: by May 20th the company had filed for bankruptcy. The ‘Chief Wizard’ of this AI Oz had been exposed. 

The ‘Sucker’

And something else had been exposed too. 

For employees and customers, the fallout was considerable, but major investors deserve scrutiny.

Builder.AI attracted significant backing from Qatar Investment Authority, Microsoft, Insight Partners and others, who led a $250 million Series D round in 2023, valuing the company at $1.3 billion. Earlier rounds, like the $29 million Series A in 2018 had also drawn huge interest. 

Yet, surely, following the WSJ investigation, things should have been clear. So, how did these investors fail to do their due diligence?

Negligence? Fear of missing out? An unwarranted faith in technology?

Hype Beasts

The investment strategy of large companies, who are awash in liquidity, is to fund any company whose stock is popular, in case one turns out to be the next big thing. Highly diverse, speculative investments, it seems, are made without much due diligence, and that has served the sector well. More cynically, it might also indicate a willingness among investors to pump a stock thinking they will know when to pull out before it deflates.

This makes it easy for genuinely innovative firms to be overshadowed by opportunists misrepresenting their capabilities and flooding the market with AI slop. AI-washing mirrors the dot-com bubble, where unproven tech companies drove valuations to unsustainable heights. 

The current frenzy then risks a similar crash, as investors and consumers grow skeptical of the highly inflated promises made by AI’s purveyors.

Statement

Builder.ai promised users would be able to ‘build custom software like ordering pizza,’ and major investors—including SoftBank and Microsoft—poured capital into its Natasha tool. Yet despite founder Sachin Dev Duggal’s impressive résumé, Natasha relied on rudimentary technology and low-wage developers rather than genuine AI. A 2019 Wall Street Journal exposé revealed this, but investors continued lavishing the company with funds. By May 2025, Builder.ai went bankrupt, a good example of AI-washing enabling opportunistic start-ups to drive a hype, resulting in the undermining of confidence in legitimate AI innovation.