Check and Prevent backdoors from security providers

Dear Experts,
i have a question. How do we determine that the security solution that is deployed, whether it be endpoint protection, AV or something else, doesnt establish a backdoor and share the sensitive information collected from the endpoints?

Like for example, for AV updates, the model is to pull from cloud and ideally no data should be sent back to cloud servers of that AV. Does firewall help to detect this pattern in terms of data payload?

collect settings, printscreen when open program, sample files suspecious… no more than that >:-D

if you use windows or any av you have nothing secret/sensitive