It's all data, that's true. But I think the distinction the various three-letter agencies make is between "metadata" and "content" for most data they process.

It's all data, that's true. But I think the distinction the various three-letter agencies make is between "metadata" and "content" for most data they process. So in terms of a phone call, an actual voice recording would be "content" while when the call was made and between which two numbers would be "metadata". With "content" + "metadata" they could verify that the target is actually who they think it is, based on voice. With just "metadata", the chances are higher that it might not actually be who they think it is.