Check with the legal fine print. Universities, colleges see their research as fair game for IP differently. At a local Uni I know of going through the IP application on using technology developed withing the area of study I found things like:

- response time for application? (4 months)

- response time for decision approval of application (within 4 months)

- commercialisation was something like 10% back to the University + 30% return for future funds (what ever that is) on profits made. (meaning there has to be submissions made back to the University on accounts).

But it will vary widely. So check. For instance did you realise, Google has a long term licensing arrangement with Stanford [0]

For me though the rub is even if you are not developing in your area of research do you want to be shackled having to submit information and asking permission?

Reference

[0] Google, PageRank, 'In the Crosshairs'

http://www.jimworld.com/gazette/issue-208/1053114758.html

I don't know about biotech, but in sw, not unless (a) it's based on your research and (b) there are patents involved. Even then universities tend to be accomodating. They make more money by not pissing off founders.

At Pitt, if you have patentable research where a startup spinoff is possible, you submit it to the tech transfer office.

Option 1: they don't think it's worth patenting, you can do whatever you want, so long as you give them 15% of revenue resulting from said idea.

Option 2: They patent it for you and you get 30% of revenue.

Either way, you really don't want the university to get hold of your IP, except you can get a free patent in your name out of it.

You should get some documentation showing whatever you developed, you developed on your own time.

Hmm, you're right.

http://www.pitt.edu/HOME/PP/policies/11/11-02-01.html

I suspect these rules are really meant for biotech research, though. They'd never get away with applying them to web stuff. If I were a student and started working on a web project, I'd ignore these rules.

Harvard's policy

http://www.techtransfer.harvard.edu/PatentPolicy.html

says, except for some medical stuff, your inventions are your own.

So Pitt's policy is in fact unusually harsh. I suspect they wouldn't dare enforce it in practice.

Pitt's policy is pretty much completely ignorable if you're a student, unless you go out of the way to lose your IP rights. There's been a campaign to get more faculty to submit "invention disclosures" to the tech transfer office but I'm not sure how aggressively the university would litigate if a faculty member just didn't submit their research and proceeded to commercialization themselves.

However, the policy is less draconian when you realize that the revenue split only applies to patent licensing revenue. A professor's start up company would not have to pay licensing fees to the university. It really only makes sense for biotech because there is almost no money in licensing new software patents.

Perhaps more interesting is that Pitt's tech transfer is run at a net loss -- the vice provost said recently that they are just now breaking even which means they are probably still losing money.

You probably signed some sort of Student Handbook at the beginning of the year that says what, if anything, the university thinks it can claim.

As others have said it also probably depends on if you are doing research when you start your company or just a student.