@futurebird should also include developing methods to break the security of computer systems which are invariably used against the innocent without a warrant and make a mockery of the fourth amendment. this is a huge and burgeoning industry which conceals itself by describing its work as "defensive".
there is a huge debate in which the operative question is reframed into whether it should be allowed to "research" by attempting to develop methods to undermine security, exacerbated by the DMCA which says no, that's a felony (this law is obviously a terrible thing). "research" is one thing but there is an entire journal publication structure where people quite seriously just post ways to hack systems without a corresponding description of the defense against it. this is a heinous set of incentives.
i am heartened by the small few cryptographic publications which solve real problems with earlier solutions, but many cryptography papers will use information theory to deanonymize people and nobody is publishing about retaining anonymity except for some very limited papers on tor and i2p.
when someone works in "infosec", they tend to be one of my favorite people or potentially some of the worst and i would like this to become more widely understood as a weapons industry.
the only jobs i have been offered for a while were for places that clarified they would involve hacking tools for three-letter agencies. i would rather die than live with myself having contributed to harm like that.