Windsurf, Stock Options and the Hard Job of Regulating

Recently, OpenAI attempted to purchase Windsurf for $3 billion. Windsurf and OpenAI wanted to keep the newly acquired technology to themselves and not share it with Microsoft.

However, Microsoft is OpenAI’s biggest investor, and as part of their partnership, Microsoft has rights to OpenAI’s core IP, including new IP acquired through acquisitions.

Microsoft and OpenAI couldn’t agree on terms that worked for both parties. There was a 60-day exclusivity window during which Windsurf couldn’t entertain any other offers.

As soon as the 60 days expired, Google swooped in, paying $2.4 billion for licensing and hiring top engineers. A few days later, Cognition bought the rest of Windsurf.

There’s a lot to unpack here, so let’s dive in.

The biggest question: why didn’t Google just buy all of Windsurf? The answer is simple—they wanted to avoid regulatory scrutiny.

Here’s the backdrop. A few years ago, the antitrust division of the DOJ(Department of Justice) filed a case against Google, accusing it of being a monopoly. The DOJ won, and Judge Amit Mehta officially declared Google a monopoly. Now, the DOJ is deciding how to break Google up, a process that typically takes years.

The antitrust crackdown isn’t just on Google. The DOJ also filed a case against Facebook (now Meta). That trial is ongoing. As we all know, Meta owns the top three social networks in the U.S.—Facebook, Instagram, and WhatsApp. Only Facebook was homegrown; the other two were acquisitions.

When Facebook bought Instagram, the DOJ didn’t raise concerns, so the deal sailed through. The same happened with WhatsApp. Now, the DOJ is looking back, realizing that too many giants—Amazon, Google, Meta—have grown unchecked, and they’re trying to correct those past mistakes.

As part of this “correction,” regulators have gotten aggressive, blocking big tech from acquiring companies outright. Remember, the DOJ recently blocked Adobe from buying Figma. That’s the regulatory environment we’re in now. Big tech is afraid to acquire outright, fearing the DOJ’s wrath.

So, what’s the workaround? Don’t buy the company—just hire the key people and license or acquire the tech.

Just last month, Meta took a 49% stake in Scale AI for $15 billion. Do you think Meta doesn’t have another $15 billion to buy the rest of the company? Of course they do. They didn’t because they knew the DOJ would challenge it.

Now, technically, the DOJ can’t outright block deals. They can sue, and a judge ultimately decides. But if the DOJ fights an acquistion then the whole deal gets delayed by one to two years, because nothing can close until the court rules.

So Meta did the next best thing: they hired Scale AI’s CEO, Alexandr Wang, a few top engineers, and got the company’s IP—without triggering a full acquisition battle.

But here’s the problem: these workarounds hurt startups and their employees. A startup needs a whole team to function—engineers, designers, QA, marketing, sales, everyone.

When a company used to get acquired, everyone benefited through stock options.

Now, that norm is breaking. Google got Windsurf’s top engineers. But what about the rest? The HR team? The designers? The QA team? They were left out. Because the company wasn’t technically acquired, they didn’t see any of Google’s $2.4 billion payout.

Luckily, Cognition acquired Windsurf later, giving the rest of the team a happy ending. But it easily could’ve gone the other way.

Look at Scale AI. Meta paid $15 billion for 49% and the top engineers. What about the rest of the employees? They still hold stock options, but those are worth nothing until another Cognition-type buyer comes along.

If this trend continues, why would a talented designer, QA engineer, or technical documentation writer join a startup? They know they’re not the ones a big company will cherry-pick. They’re better off taking a stable job elsewhere.

Working at a startup is already risky. The upside—the potential payout when the company gets acquired—is now less certain. In this new regulatory climate, the biggest beneficiaries are big tech.

When big tech used to acquire a whole company, they’d inherit all the staff—marketing, QA, designers—whether they wanted them or not. Maybe after six or nine months, they’d lay some off. But those people at least got to work at Google or Meta, collect salaries, and walk away with high severance.

Now? Big tech can just pick the engineers they want and skip the hassle. Since they are hiring folks there is no delay. Engineers can join in a few weeks.

So, who’s to blame for this new environment? Should the DOJ loosen up? Well, it’s complicated.

Remember, the DOJ blocked Adobe’s acquisition of Figma. Now Figma is heading for an IPO—a fantastic outcome for competition and the industry. If Adobe had swallowed Figma, innovation would’ve suffered. So there are real benefits to the DOJ’s tough stance.

And yes, Google is a monopoly in search, and their results have gotten worse because they face no competition. The DOJ was right to file that case.

My point is good efforts often have collateral damage, and that’s what we’re seeing. Right now, the DOJ is trying to make up for decades of lax oversight. During this tightening phase, we’ll see more cases where companies scoop up key people and IP instead of acquiring the whole company.

Unfortunately, in this climate, working at a startup with stock options feels riskier than ever. If you’re a smart person who’s not a founder or founding engineer, working at a big company might be the safer bet — for now.