In my view, unfortunately the answer is: unless you're willing to go with the two pretty radical options I mention below, probably not with any kind of really strong guarantee of security.
Other answerers have suggested technical measures that, in theory, might possibly allow users to see information on screens properly while preventing cameras from doing so. With all respect to those views, frankly, I'm a little skeptical that there are any such measures that would reliably work, at least if we're talking about facing an opponent actor with some minimal ability to do things like use photo and video editing software to recover any information from pictures/frames that might indeed be successfully made difficult to see on ordinary viewing. At the least, I would not rely on any techniques like that to meet the security need you're talking about without first having some extensive, independent testing in-hand demonstrating strong effectiveness. Which I dount you're going to be able to find.
If we limit ourselves to measures that we know will work with a high degree of likelihood if implemented properly, unfortunately we are left with two admittedly not-great options:
Implement tight, physical security searches to prevent any employees from bringing any kind of electronic devices into a highly-secured area where the computer screens thatshow the sensitive info are kept.
Rework or replace the software that displays the sensitive data on screen (or the ways that you use that software) so that, well, the data is never actually on screen.
Option 1 is how governments and enterprises secure super-sensitive information that they must protect in high-security facilities. It is difficult, and often-times expensive, to implement. (You need, at the very least, dedicated security personnel screening each person who enters the secure area.) Option 2 is more palatable in many ways, except that depending on your workflow and how your workers need to do their jobs they might well need to see the actual clear-text sensitive information on screen. Whether that's a practical course or not depends on how your business or organization actually uses sensitive information.
Now, all of the above being said that doesn't mean that there aren't measures you can take to reduce the risk of an employee deciding to whip out his or her phone and take a photo of on-screen information. Obviously, you can impose a policy ban on bringing devices with cameras in the areas where computers with sensitive info on them are located, and let your employees know that if they are caught breaking that rule punishment will be significant. And of course you can and should do background checks on anyone before allowing them access to sensitive information in the first place. But those policy-based measures are, obviously, very far from foolproof.
In sum, the taking-a-picture-of-a-screen scenario is just a really rough one to combat. If you can prevent sensitive data from ever really being on screens to begin with that's probably your best of a bad set of possible options if you aim to very strongly protect the confidentiality of it.