Builder.ai: When AI Washing, Fraud Leads to Collapse - and D&O Insurance Doesn’t Catch You

By Steven Barge-Siever, Esq.

Founder | Upward Risk Management LLC

Collapse of Builder.ai

Not long ago, Builder.ai was a $1.5 billion AI unicorn, backed by Microsoft and endorsed as the future of automated software development.

Today, Builder.ai is insolvent.

I remember seeing their team at Silicon Slopes in Utah (2023) - a full entourage in matching shirts, radiating confidence. I was impressed by the show - though a bit puzzled when I learned it cost $25,000+ to build something relatively simple. I didn't really understand why it would be so expensive for “automated” software....

Today, I'm still impressed - not by their presence, but by how far a company can go on AI-washing and (alleged) financial fraud. $450 million raised. Almost none of it built to last.

Hundreds of engineers were reportedly doing the “automated” work manually. And a set of transactions with VerSe Innovation is now under scrutiny for what appears to be “round-tripping” - a scheme in which two companies invoice each other for similar amounts with no real services exchanged, inflating reported revenue without underlying value.

And whatever happens next - lawsuits, investigations, finger-pointing - one question will sit beneath it all:

Will the D&O insurance respond?

As a broker and attorney who’s advised on complex risk for 15+ years, I’ve seen how these situations unfold, and I do love relevant examples that illustrates problems that can arise with insurance.

So let’s walk through the story and what’s likely to happen. What claims are coming. How D&O responds. Where it fails. And how we would’ve structured the protection if we had been there before the collapse.

The Rise: How It Looked on the Way Up

Builder.ai sold an irresistible vision: an AI-powered platform that could build software automatically - no developers, no code. Just an interface, a business idea, and instant software.

That narrative unlocked:

• Nine-figure funding rounds

• Strategic partnerships with Microsoft

• Valuations north of $1.5 billion

• Board and investor confidence

  
  

But behind the scenes, the AI wasn’t doing the building. Reports now show hundreds of engineers - manually coding what the platform claimed to automate. And revenue? Allegedly inflated through round-tripping transactions (moving money or services in a circular manner with no real economic substance) with VerSe Innovation.

The Cracks Became Visible

Builder.ai’s collapse wasn’t entirely unpredictable. In fact, the signs were there as early as 2019.

A Wall Street Journal investigation found that despite bold marketing claims, most of Builder.ai’s software was being built manually. The AI? Mostly smoke and branding.

In 2020, former executive Robert Holdheim sued Builder.ai for $5 million, alleging he was fired after flagging deceptive investor communications. Legal filings revealed that the company had claimed apps were “80% built” by AI - even though the underlying tech barely functioned.

Other ex-employees echoed the reality: “All engineer, no AI.”

By early 2025, the internal rot had reached the balance sheet. After a leadership shake-up, new CEO Manpreet Ratia allegedly discovered that Builder.ai had reported $220 million in 2024 revenue when actual income was closer to $50 million (source).

The illusion of automated scale had finally hit a wall. And once the story unraveled, the capital, the confidence, and the company itself quickly followed.

From my POV, this wasn’t just operational mismanagement - it was a disconnect between the narrative and the truth.

And in my world, we call that Fraud.

As for D&O? Fraud is where the real problems begin.

A Quick Primer: What D&O Insurance Actually Covers

Before we dive deeper, it’s worth pausing to explain how D&O insurance is structured.

D&O (Directors & Officers) insurance isn’t one blanket policy. It’s structured into three parts - Side A, Side B, and Side C - plus an optional fourth: Side A DIC. Each serves a different purpose depending on who is being sued, who is paying, and whether the company is still standing.

Here’s the breakdown:

Side A - Personal Asset Protection

This is what protects individual directors and officers when the company can’t or won’t indemnify them.

• If the company is bankrupt (The case at hand)

• If indemnification is legally barred

• If the board member is named in a derivative suit

Then Side A steps in and pays that person’s defense and settlement costs directly.

In collapse scenarios - like Builder.ai - this becomes the most important layer of protection.

Side B - Entity Reimbursement

Side B kicks in when the company does indemnify its directors and officers - and wants to be reimbursed.

For example, if Builder.ai paid a board member’s legal fees and settled a claim, Side B would pay the company back. But it doesn’t help if the company can’t front the money - or is out of business.

Side C - Entity Coverage

For private companies, Side C covers the company itself when it's named in a claim.

So if Builder.ai is sued directly (e.g., by investors or regulators), Side C covers the company's legal costs.

However, there’s a catch: Side C shares limits with Side A and B. That means when the company defends itself, it’s draining the very same limits individual directors may later need.

Side A DIC – The Backup Plan That Actually Pays

Side A Difference-in-Conditions (DIC) is a separate policy - a sort of "escape hatch" for directors when the base policy fails.

It’s:

• Non-rescindable

• Not shared with the company

• Triggered when coverage is denied, rescinded, or exhausted

It exists to protect individuals in exactly the kind of situation Builder.ai now faces - where insolvency, fraud allegations, and internal disputes converge.

The catch? Startups don’t buy it. They treat it like an add-on - until it’s the only thing that could’ve saved them.

This is the framework behind every (private company) D&O policy. And whether or not it works when the crisis hits? That comes down to what was negotiated before everything fell apart.

The Unraveling: What Claims and Litigation Lies Ahead?

1. Investor Lawsuits for Misrepresentation

Shareholders and late-round investors will claim they were misled - that Builder.ai’s AI claims, revenue figures, or growth forecasts were materially false. They will undoubtedly seek reimbursement from the now insolvent company. They will likely also go after perceived bad actors that profited (directors and officers), and potentially their personal assets.

And the costs of defending and settling this litigation process is what D&O insurance is designed to cover - and as you will see not cover.

D&O Response: Side A (Personal asset protection), Side B (indemnification of individuals) and Side C (entity-level securities coverage) are designed to cover this.

But coverage stops if fraud is proven - or if the insurer rescinds the policy.

What We Would Have Negotiated:

• A fraud exclusion that only kicks in after a final, non-appealable adjudication - no denial based on settlements or SEC decrees

• Full severability, so one exec’s misconduct doesn’t wipe out the board’s protection

  • I.e. one bad actor does not invalidate coverage for those unaware of the alleged fraud

    
    

2. Regulatory Investigations: FCA, SEC, FTC

When bold AI marketing and inflated numbers meet consumer and investor protection rules, regulatory agencies show up.

Expect:

• FCA or SEC investigations into financial misstatements

• FTC scrutiny under deceptive trade practices

• Possible coordination across jurisdictions

  
  

D&O Response: Covered for individual defense, but typically subject to sublimits - often $250K to $1M - and only triggered by formal proceedings.

What We Would’ve Negotiated:

• A broader definition of “Claim” to include subpoenas and informal inquiries

• A full or higher regulatory coverage

• Advancement of defense costs tied to retention of counsel, not regulator formality

  
  

More in: Startup D&O and SEC Pitch Deck Risk

What Happens Next? Likely a Battle Over Side A

Once the lawsuits hit, Side A becomes the most valuable part of the D&O policy because there’s no longer a company to indemnify the individuals (Side B).

But even Side A isn’t bulletproof.

Here are the complications:

Who controls the policy?

1. Most policies have a Priority of Payments clause - deciding who gets paid first and when:

   1. Side A (unindemnified individuals)

   2. Side B (indemnified individuals)

   3. Side C (the entity)

If this isn’t clearly structured, the first claimant - or the person with the loudest legal team - may drain the policy.

What to Negotiate:

• Priority of Payments favoring Side A first

• Contractual notice before limits are exhausted

Did they purchase Side A DIC?

Standalone Side A Difference-in-Conditions (DIC) policies exist for exactly this kind of collapse. They are:

• Non-rescindable

• Excess over the base policy

• Designed to protect individual directors when the entity fails

But many startups skip them.

Will the insurer cancel or rescind - Impact of Application Misrepresentation

Insurers analyze D&O risk, and assign premiums/pricing, based on the information contained in an application - a signed statement, often by the CEO, CFO or GC, certifying that material facts are accurate.

If Builder.ai lied about financials, described itself as “AI-native” or “automated” in that application the insurer may allege material misrepresentation and seek rescission or cancellation.

• Rescission: Voids the policy from inception

• Cancellation: Terminates the policy mid-period (in some cases)

Most D&O policies allow cancellation only for non-payment, but some allow broader triggers.

What to Negotiate:

• Cancellation rights limited to non-payment of premium

• Runoff coverage upon change in control or insolvency

What about the fraud exclusion?

A poorly written fraud exclusion lets the insurer deny coverage "for fraud" - when alleged or if “established in fact,” even without a final verdict.

What to Negotiate:

• Final, non-appealable adjudication trigger in the underlying matter

• No clawback of defense unless guilt is proven in court

So this isn’t just a question of whether D&O responds. It’s a question of:

• Who keeps coverage

• Who loses it

• How fast the limits are consumed

• And whether there’s anything left for those who didn’t know what was going on

A Turning Point for Underwriting

Builder.ai didn’t collapse because the product failed. It collapsed because the story failed — and it took a $1.5 billion valuation down with it.

That changes things for underwriters.

We’re now in an era where:

• AI claims must be vetted, not just accepted

• CFO and CTO application signatures matter

• Insurers will ask what’s truly automated - and what’s just manual labor with a brand on top

We covered this in AI Washing: The Dirty Truth - and now we have the cautionary tale to go with it.

The Real Test of D&O Is When the Story Falls Apart

D&O insurance is easy to overlook when valuations are rising and press releases are glowing.

It’s also easy to treat the application as a formality, the fraud exclusion as boilerplate, and the terms as “market standard.”

But when the narrative breaks - when investors feel misled, regulators show up, and employees start whispering to reporters - D&O becomes the last line of defense.

Not just for the company. For the people who signed off on the funding, the pitch, the audits, and the promises.

The Builder.ai collapse won’t be the last of its kind. But it should be a lesson to anyone signing D&O applications in a market that rewards hype - especially in AI, where product and marketing can drift dangerously apart.

If you’re building or investing in companies like this, make sure the insurance does more than exist.

Final Thought

Getting D&O right doesn't have to cost more money or take more time - it just needs to be negotiated by a broker or attorney that understands the implications of placing it poorly.