I understand that it is impossible to prove a negative, but at the same time given the scope of the surveillance dragnet I'd hope they'd have something they could point to. At this point I would even circumstantial evidence as to the programs efficacy would go a long way, but it doesn't look like they even have that.

Its not proving a negative, its proving an existential query, there exists terror attacks that were stopped by PRISM. If the data was useful, then they could certainly point to the in-progress attacks that were thwarted - stopping an attack does not erase all information about its existence.

The government prevents thousands of terrorist attacks every year by being tolerable enough that the number of people filling rental trucks full of explosives is zero.

? There are actually a myriad of interlocking protocols that help to thwart such activities, to ignore them is not reasonable.

Most poignantly there is literally an Ammonium Nitrate security program. [1]

As for the 'number of truck bombings being zero' ... it did happen and the average is 'more than zero' and when we consider the implications (an entire building and 100's of dead etc.) it's a concern.

Obviously, it's hard to determine the effectiveness of the various programs because they all play a part, but that there haven't been any such bombings since then is at likely a positive sign.

[1] https://www.cisa.gov/ammonium-nitrate-security-program

Timothy McVeigh begs to differ (https://www.theguardian.com/us-news/2015/apr/13/oklahoma-cit...)

That was 25 years ago. The average number of such trucks per year is zero.

The "secretly prevented attack" thing is bunk because really prevented attacks, and the plotters of those attacks, end up in court and in prison.

Do they? What about attacks prevented before a crime has been committed?

By your logic, gun control can’t prevent any shootings.

How do you know it was only one?

Because he prevented it!

Sure but he could have prevented others and not know about it!