You are never allowed to steal trade secrets or use them elsewhere even if you didn’t sign any agreement about it.
An NDA isn’t even technically required. There are state and national laws which blanket ban using trade secrets outside of the company you worked for.
Not saying anything is right or wrong, just that when something is illegal doesn't mean it prevents that crime.
They don't. What I learned while working for your company isn't your company's property any more. Good ideas almost always eventually spread in our industry. And I think thats a good thing for software as a whole!
You can protect your data, and your code. But you can't really stop someone quitting a job at your company, working somewhere else and reimplementing a software system that worked well. It might take years to do it, but probably not decades.
I honestly think this is a pretty good tradeoff. It means if you build some software, you have head start, but not an impenetrable wall. For someone to compete, it'll take a lot of time and money just to catch up with where you are today. So it'll be hard to do but possible. This leaves the door open for any incumbent to be outcompeted in the market if they stop doing good work.
And thats a good thing! Competition is painful, but it pushes us to make better products for our users. Ultimately thats better for everyone.
I don't even think this case is undesirable. If we were welders, it would be absurd to be prevented from using a welding technique we learned on the job at a new employer. System design is just a technique.
You are only thinking about code. Imagine knowing all the dirty secrets about how your company screwed customers. I have seen employment contracts forbidding working for a customer, in addition to competitors and in addition to an NDA.
Where trade secrets really leak (in tech), in my experience, is engineer to engineer. A chat between friends over beers about some technical problem. It is nearly untraceable and it doesn’t involve anyone leaving their job.
As a fun example, database tech is buried in trade secret restrictions and has been for decades. There is a classic problem in cache replacement algorithms that has no solution in literature. Nonetheless, an astonishingly elegant solution exists — the kind that you can’t believe you never thought of it yourself after you learn it — that has been selectively passed around informally among practitioners for (at least) a decade or two. No one knows who invented it but it was likely developed at one of the old database research powerhouses like Oracle, IBM, et al that have severe trade secret regimes. A trade secret that leaks isn’t a trade secret, but there are enormous punitive consequences if anyone knows who leaked it.
This kind of trade secret leakage happens even under non-compete regimes and it is pretty common. When it happens, the probability of figuring out how it happened is very low. It has to be part of your risk model.
A non-compete agreement doesn't actually "prevent" an employee from working at another company any more than an NDA prevents an employee from divulging confidential trade secrets.
If you're talking about trade secrets, I believe if there is evidence in the product/products that a former employees company is releasing that seems to be operating or working in a similar product they could gather publicly available evidence, hire a PI, and ultimately attempt to subpoena additional information if there is sufficient evidence that the employee is actually sharing trade secrets from a previous company.
> Anthony Scott Levandowski pleaded guilty and was sentenced today to 18 months in prison for trade secret theft related to Google’s self-driving car program, announced United States Attorney David L. Anderson and John F. Bennett, Special Agent in Charge of the Federal Bureau of Investigation. Levandowski was also ordered to pay a $95,000 fine and $756,499.22 in restitution.
