In 2002, Harvard students Cameron Winklevoss, Tyler Winklevoss and Divya Narendra began developing an idea for a Harvard-only social network, first called HarvardConnection and later ConnectU. Early in the project, they recruited a fellow Harvard student to help write the site's code: Mark Zuckerberg.
What happened next is genuinely disputed, and it has stayed disputed for two decades. According to Winklevoss and Narendra's account, Zuckerberg agreed to help build their site, then delayed for weeks while secretly building a separate social network of his own. He launched that site, TheFacebook, in February 2004. Facebook and Zuckerberg have disputed this characterization and maintained he did not take ConnectU's code or its idea. (The 2010 film "The Social Network" dramatizes a version of this story; it is a film, not a factual record, and should not be treated as one.)
ConnectU sued, alleging Zuckerberg had taken their concept, and in some accounts elements of their code and business plan, while working for them. Facebook and Zuckerberg countersued, alleging ConnectU had improperly accessed Facebook's computer systems.
There was no agreement of any kind between the two groups covering what Zuckerberg was building, who would own it, or what he owed ConnectU if their project failed while his succeeded, only an informal understanding that he would help build a site. That absence of anything in writing is a large part of why the case became a multi-year fight over intent and credibility, rather than a five-minute reference to a contract.
The parties settled in 2008, for a package reported at $65 million: $20 million in cash and $45 million in Facebook stock, valued at the time of the deal. The Winklevosses later tried to reopen the settlement, arguing Facebook had misrepresented the stock's value during negotiations. The Ninth Circuit Court of Appeals rejected that challenge in April 2011.
At some point, litigation must come to an end. That point has now been reached.
That line, from Judge Alex Kozinski's opinion for the Ninth Circuit, effectively closed the case. The Winklevosses chose not to pursue the matter further. In the years that followed, as Facebook's stock rose, the stock portion of a settlement they had once tried to unwind as too small was reported to be worth well over $160 million.
What founders can learn from this
Whatever the full truth of who owed what to whom, and after this many years, it is still argued about, the case sits on a gap that shows up constantly in early-stage teams: someone is recruited to help build a concept before any company, or any agreement, exists. In that gap, there is no document defining who owns the output, what "helping" entitles someone to, or what happens if the relationship ends badly. Who owns the IP before incorporation is written for exactly this moment: before a company exists, whatever gets built belongs, by default, to whoever built it, not to whoever had the original idea, unless everyone involved puts something in writing early that says otherwise.
The practical lesson is not about predicting who would be right in a dispute like this one. It is that the moment you ask someone, even a friend, even a fellow student, to help build something you consider yours, is the moment a short written understanding, about roles, about ownership, about what happens if it does not work out, stops being optional. It is far cheaper to write while everyone is still on the same side.
Sources: CNBC, Bloomberg, The Register, FindLaw.