The Rise of the IP Licensing Deal
Google’s Windsurf deal and the new M&A playbook reshaping tech
Google’s deal with Windsurf, a popular AI-powered coding assistant, dominated the tech news this weekend. The deal follows a string of similar IP licensing deals over the last year that have taken place between large tech companies and emerging AI startups.
Many in tech have taken to Twitter/X to decry the deal, and what they view as an unfair outcome for the employees who may not be paid as part of the stock ownership they’ve earned in the company. I think this narrative is overblown - the employees will probably get a great payout commensurate to their vested equity ownership, and I expect that part of the story to come out in the coming days. I believe this because the opposite scenario, leaving their employees with nothing, would be an embarrassing abdication of leadership on the part of the management team.
Still, there is a strong case to be made that these IP licensing deals are not good for the tech ecosystem. Like them or not, they are likely here to stay, so anyone who works in the startup ecosystem - founders, employees, investors, LPs - needs to understand how these deals work.
There has been a lot written about the details of the deal, so rather than provide an exhaustive re-hashing, I’ll share how these deals work, the incentives on both sides of the table, explore why they are so controversial, and lay out a path forward for the industry.
The high-level details of the deal
Google’s announcement to strike an IP licensing deal with Windsurf comes after months of OpenAI being rumored to acquire the company. The OpenAI-Windsurf talks fell through because of OpenAI’s partnership with Microsoft, which gives Microsoft access to all of OpenAI’s technology. The competitive dynamic between Windsurf and Microsoft’s suite of software development tools, in particular GitHub, created angst on both sides that ultimately killed the deal.
Enter Google. The company announced it would be paying $2.4B to Windsurf to license its technology, hire the founders and key leaders, and pay out investors and employees. Employees who have unvested stock will not be receiving a payout, according to the most recent reporting.
Big tech’s new acquisition playbook
The Google-Windsurf deal follows a similar playbook that Google executed in their $2.7B deal with Character AI. Microsoft had a similar deal with Inflection for $650M. Amazon had deals with Adept AI for $1B and Covariant for $1B. And more recently, Meta just announced a similar deal with Scale AI for $14.3B.
There are nuances to each of these deals, but they represent a similar flavor of IP licensing agreement that has become the norm among big tech companies. They share a few traits in common:
Startup maintains independence. The buyer doesn’t take any equity stake in the startup. The startup remains as an ongoing company with employees, shareholders, and a product it sells to its customers. All of the equity in the startup becomes owned by the employees, and the investors relinquish their equity in exchange for cash.
Non-exclusive licensing. The buyer gets access to the core IP and technology from the startup on a non-exclusive basis. Non-exclusivity means the startup technically could be free to strike similar types of deals with other companies in the future. In other words, the startup still owns its product, and is free to do whatever it wants with it.
Senior talent acquisition. The buyer hires key personnel, often the founders and senior technical leaders, to join as full-time employees. The startup appoints senior leaders who stay behind into C-suite positions, who can lead the company in whatever direction they see fit.
Cash payouts. The buyer pays a lump sum to the startup, split across three key recipients:
1) The founders and senior leaders who will go work at the buyer’s company,
2) The investors in the startup who will receive cash in exchange for giving up the equity in their company
3) The startup itself, which will retain that cash on its balance sheet and use it to fund future operations. That includes a dividend to shareholders, which are now the employees.
So, to sum it up in the case of the Google-Windsurf deal: Windsurf will remain an independent company, wholly owned by the employees who stay. Venture capitalists who funded Windsurf, such as Kleiner Perkins, will get a return on their investment and then no longer own stock in the company. Windsurf’s product will be licensed to Google, and Google will be able to build their own version of it to sell to their customers.
Dual incentives
Big tech’s M&A playbook is increasingly relying on these IP licensing deals over traditional acquisitions because they prioritize speed by evading regulatory scrutiny. A traditional acquisition is subject to government review by anti-trust regulators, who assess a deal’s potential to create a monopoly in the merged companies. If the regulators deem a merger or acquisition would create too much market power, they can block the deal.
IP licensing deals on the other hand, are structured as commercial agreements, with the startup continuing to operate as an independent company post-transaction. So, they are not acquisitions (wink wink).
In reality, these deals are absolutely acquisitions.
In an era of $100M salaries to top AI employees, the scarcest asset in the tech industry is talent. With these IP licensing deals, big tech gets access to a startup’s top employees and their core IP - without needing to deal with the headache of integrating dozens or hundreds of people.
While it may seem like the incentives are strong for the buyer, there are also strong incentives for the startup. The incentives are strongest for the founders and investors. They both get a meaningful cash payment, and in the case of the founders, a life-changing sum of money plus the opportunity to become a leader at a big tech company. They get to run the product they previously had built at their startup, but do it with the seemingly-limitless resources of a global tech company. For the investors, earning something between a good and great return on their investment, in a relatively short time period can also be a great deal. Especially if the company is number two in its category, the way Windsurf was to Cursor, the leader in the code generation space.
Where the incentives are weaker are for the startup’s employees. Depending on how the startup negotiates its terms, employees may or may not receive payouts for the stock they’ve been paid as part of their compensation.
Why these deals are so controversial
Silicon Valley and the broader tech ecosystem rely on trust. Since Arthur Rock wrote his Intel memo in 1968, the standard employee compensation package in the tech industry has included equity that creates shared upside for employees and management in the event of a successful exit.
In an acquisition - whether traditional or these IP licensing deals - it’s incumbent on the CEO of the startup to ensure their people are being taken care of in the deal. It’s unclear at this point to what degree the Windsurf employees are being taken care of, but we will likely find out in the coming days.
For many people though, the sight of a startup CEO decamping to a big tech company with a huge bonus and leaving the startup behind to go it alone, is the opposite of Silicon Valley’s in-this-together ethos.
In IP licensing deals, the startup’s core IP gets diluted because the big tech company can sell it to their customers. If Google’s customers can buy Windsurf from Google, bundled into their Google Cloud bill, why would anyone buy Windsurf from Windsurf? Plus, a startup without its founders can become strategically rudderless, especially before it has scaled to become a profitable and sustainable business. The morale hit from all of the above can kill a startup’s culture, and its ability to attract and retain talent.
So why don’t the startup’s founders just pay out their employees and shut down the company? The answer here is murky, but my guess is because claiming that the startup is going to continue on as an independent entity provides cover from government anti-trust scrutiny.
How tech can move forward
We are likely to see more of these IP licensing deals in the future given how rapidly AI is upending the competitive dynamics in the tech industry. The tech industry needs a plan moving forward.
Understand why this is happening. AI - it poses both the greatest existential risk and the biggest value-creation opportunity for every big tech company. This dynamic is compelling them to act aggressively and decisively to advance their business interests and stave off competitive threats. In the era of AI, big tech companies need two things - speed, and AI talent. Structuring deals as IP licensing agreements rather than traditional acquisitions solves for both, as big tech gets the AI talent without the government roadblocks.
Don’t blame the government. Many people are reflexively blaming the government’s anti-trust policy, which in some cases like Adobe/Figma as creating incentives for companies to skirt the M&A process with these IP-licensing deals. This misses the bigger picture. Yes, anti-trust policy has been unnecessarily heavy-handed in some cases, and the rhetoric from both political parties against big tech has worsened the problem. The fact is these deals are happening because of big tech’s need for speed + AI. Robust anti-trust policy is a fundamental part of a healthy capitalist economy. Big tech should engage policymakers to update antitrust laws to allow for rapid anti-trust resolution in emerging categories like code gen.
CEOs need to be held accountable to take care of their people. In any acquisition, the CEO has ways of ensuring their employees get compensated for their equity, or not. These IP licensing deals do not implicitly favor the management team over the employees any more than a typical acquisition. It is incumbent on the CEO to take care of their people, and in cases where they don’t, they deserve the public criticism that they’ll receive. In the case of Windsurf, the CEO has taken a lot of barbs on Twitter/X, but it’s probably too soon to tell whether it’s actually deserved or not.
New compensation structures. Startup employees should push for compensation structures that result in fair payouts in the event of an IP-licensing deal. These can include acceleration clauses or "change of control" provisions that trigger immediate vesting or cash equivalents for equity holders in an IP licensing or talent-acquisition scenario. Fair outcomes for employees shouldn’t rely on the threat of public scrutiny of a CEO’s actions in these deals.
These IP licensing deals are likely to become more common, as they enable big tech companies to move quickly, hire top-tier AI talent, and avoid getting blocked by regulators. That is, until the government recognizes that these deals are acquisitions and starts looking into them. In the meantime, IP licensing deals risk undermining the core incentives of shared success that have made Silicon Valley attract ambitious talent willing to take the risks inherent in early-stage startups. Protecting that is more than just about fairness, it’s about protecting the innovative edge that moves the tech industry forward.