Check your institution's policies. Here are three examples:
University of Auckland:
Data may not be copied in any format for sale or any other commercial purpose, nor may it be copied or reproduced for the benefit of any third party external to the University of Auckland.
So assuming the person you are sending the file to is not affiliated with the University of Auckland, you may not send the file.
National University of Singapore:
The following limits on the copying or reproduction of any Work for the purpose of research or study must be strictly observed:
Not more than, in total,
- 10% of the total number of bytes in the Work, or
- 10% of the total number of words in the Work or, where it is not practical to use the
total number of words as a measure, 10% of the contents of the Work, or
- One chapter of the Work, where it exceeds any of the limits.
So you may not send the full file.
University College London
You MUST NOT:
- provide print or electronic copies of material to anyone;
Read strictly, this is more stringent than the University of Auckland's policies; however presumably if the person you are sending to is also affiliated with UCL, then they can use their own access to get the paper. I'd guess then that UCL presumably meant the same thing as UoA, you can still send to someone affiliated with the university but not anyone external to it.
Exception: if you are one of the authors of the manuscript, you typically have more leeway with sharing. IEEE's policies are vague (see Section 8.1.4, page 76):
In return for the transfer of authors’ copyrights, the IEEE shall grant authors and their employer’s permission to make copies and otherwise reuse the accepted versions of their articles under terms approved by the Board of Directors.
But they don't say what those terms are, and I couldn't find them in the document. If you want an authoritative answer then you should email them at [email protected] and ask.
However if we use Elsevier's policies as a proxy (I'm using Elsevier because it's so big), then:
If you are an author, you may also share your Published Journal Article (PJA) privately with known students or colleagues for their personal use
So you are free to send the paper to this person, if they are your colleague in research.
If IEEE sent you a copy of the paper after they published it (the publishers I worked with did), the email might also detail their sharing policies and/or where to find them.