There are two problems here.
One is researchers who, in the past, have had collaborations as co-authors, for instance, that would today be construed as problematic. How do we get them to atone for the sins of their past?
The other is that there may still be reasonable grounds on which to have research collaborations with certain problematic actors. For instance, that might be, on the one hand, in climate change technology, but on the other hand in intelligence or military technology in terms of dual-use components. Computing is a good example.
For those, the researcher should be allowed to submit a research security plan that demonstrates appropriate risk mitigation measures, because we can't just tell a researcher, “Because you worked with problematic Chinese research or an entity in the past, you can never get any research funding again for the area in which you specialize, or you can never work with those actors again.” We should leave it up to the researcher to demonstrate that those collaborations do not pose a risk to the way Canada has articulated its red lines when it comes to its national interests and its security interests.