Certainly Google doesn't "like" to make any kind of solutions to help a dying industry. The content recognition system likely cost in the tens of millions in development and is probably eating up lots and lots of processing power just to operate it.
They have been very outspoken in their support for the "fight against piracy", and have regularly implemented "solutions" without any legal obligation to do so. One example being filtering "piracy related terms" from it's autocomplete feature in Google Search.
Google actively supports that "dying industry", because certain parts of Google, like YouTube, have a vested interest in good relations with said industry.
To me, content filtering seems like a nice bullet point in the "we don't condone copyright infringement" department. It reduces complaints that result in lawsuits, and doesn't really hurt anyone. If anything, it discourages reuse of RIAA/MPAA content and encourages the creation of new free-as-in-speech content, and that's definitely a good thing. Your cat video doesn't need a top-40 soundtrack to be enjoyable!
(It's similar to how Microsoft's attempt to lock up the web with proprietary technology resulted in Free Software web browsers becoming usable. Remember the days of "This site only works in IE"? Now pretty much every site works in pretty much any browser. Epic win.)
Which makes more sense, youtube implementing a content matching system and running it on all of the videos uploaded or Viacom and every other copyright holder making their own systems (or buying into one) that tries to go through every video uploaded to youtube?
Google's responsibility is to respond to DMCA notices; that's it. They have no requirement to create a Draconian copyright monitor.
(The standard response is "so the content industry would work with us!!!" Who gives a shit? Oh no, we wouldn't have any of the shitty "Vevo" channels laced with advertising without Content ID. What a terrible loss that would be.)