Even if the driver sent data in a format which is easily inspectable by user, and with only benign-looking data, it would be really hard to prove that something nefarious and underhanded is not being sent instead under that guise.
How is that a problem? Nobody is reviewing open-source driver code aside from the people actually working on improving it.
Also lmao; do you really believe Intel is going to sabotage their own platform? If you're doing something so "nefarious and underhanded" on your computer for that to be a concern, you likely have no reason to be concerned since you should be using a platform and OS that makes this a non-issue. Only bad-actors would be concerned and raise this as a point.
Not only bad actors would be concerned. There is a rather small but vocal minority that values privacy extremely high. They make noise, even when a threat of identification, or of siphoning out any consequential information, is extremely low.
To avoid sabotaging their own platform, Intel should remember about that, and give no food for bad PR.
As they write in employment agreements, you should avoid any real conflicts of interests, and also any appearance of conflicts of interests, because if it gets reported in media, but you end up completely clean and conflict-free, the hard-to-repair damage to public perception already would have been done.
Yeah there might be one person discussing this in social media that legitimately values privacy.
If you value privacy, you aren't using Windows. If you value privacy, you should realize that most people only care about it to the extent of it just working, and it only does so for most people running major Windows and macOS desktop operating systems If you value privacy, then you know how to read Microsoft's privacy policy to know what your data is going towards on Windows. And if you value privacy, you can respect why Intel is doing this and also read their privacy policy as to how they use the data.
People are fear-mongering that Intel is taking usage data as if Intel has some interest in using the data maliciously. None of these major companies increasing and gathering telemetry have used the data in a truly malicious manner, unless you twist the wording for specific situations where bad-actors got caught doing bad-actor activity and have to comply with legal requests for information.
This is only beneficial for the audience Intel is targeting.
> If you see where the packets are going you might be able to block at the DNS level.
There are many DNS blocklists out there and some of them do indeed block domain names from known telemetry (aka "spying") services. But if some telemetry directly report to specific IP addresses then blocking at the DNS level won't help.
FWIW my firewall blocks some companies' known IP CIDR blocks (like FB/Meta) and my DNS (unbound) blocks all known malware, porn and telemetry services (among others).
In addition to that the only process allowed to emit ongoing traffic to port 443 is the process running my browser.
Unless they circumvent the system set DNS and use a hard coded one, which many that collect telemetry seem to do. On Android I see lots of traffic to Google's DNS even though I do not use their DNS servers (and it is in fact blocked).
Even if the driver sent data in a format which is easily inspectable by user, and with only benign-looking data, it would be really hard to prove that something nefarious and underhanded is not being sent instead under that guise.