This post has been de-listed
It is no longer included in search results and normal feeds (front page, hot posts, subreddit posts, etc). It remains visible only via the author's post history.
After the "recent" fiasco in Apex Legends that turned out to be a false flag and just a regular exploit in the game servers and not in the anti-cheat I've been thinking. If Kernel Level anti-cheat can be turned on it's head and used like this and we know that's a possibility then why even do it?
There a couple groups online defending that it is unnecessary to have a anti-cheat running at Kernel Level and it is a security risk to do so and I tend to believe them.
The closest justification to use it is that it's easier to detect and that it works but ... that's not really good enough is it? Like if we can do better why don't we?
We know the internet is full of very capable people that don't need more reason than "Well ... it's there so why not?" and we know it's a possibility so at some point somebody is going to do it so, as the title says, why do we use Kernel Level anti-cheat?
There's anti-cheat tech at every level that developers are able to access, because it's an arms race against cheat developers.
Subreddit
Post Details
- Posted
- 7 months ago
- Reddit URL
- View post on reddit.com
- External URL
- reddit.com/r/TooAfraidTo...