A market agency claiming they do something of the sort isn’t proof that conversations are being monitored en masse. Security researchers can and probably have tested for this and found no clear, verifiable evidence, otherwise we would have known. Also, this stuff can be blocked at the OS level and I find it hard to imagine (esp. without solid proof) that Google or Apple would jeopardize their reputations to this extent by enabling such unauthorized listening in on users’ conversations.

Of course it’s good to keep watching this space but we shouldn’t jump to conclusions.