[Nullinker]

I would actually argue that once the masses are aware about certain technology existing and being in widespread use it becomes much easier to convince someone that a particular instance that the data is not trustworthy, so the ability to detect it through technological means becomes less important.

In the stage before widespread use people are much more easily tricked because they are unaware that others have certain capabilities which they never experienced first hand.

[nick238]

You're missing the flip side: falsely believing something is forged.

Now that the technology is so accessible and widespread, someone could deny truth by saying whatever audio/visual evidence was deepfaked, and people will believe it.