All information we collect about the use of Google Fiber TV (including use of programs and applications available through Google Fiber TV) may be associated with the Google Account being used for Google Fiber TV.
... information from the use of Google Fiber Internet (such as URLs of websites visited or content of communications) will not be associated with the Google Account you use for Fiber, except with your consent or to meet any applicable law, regulation, legal process or enforceable governmental request.
They don't really mention what information not associated to a Google account is mined except that it follows the company-wide privacy policy. So they probably mine it. It's probably comparable to Google DNS: https://developers.google.com/speed/public-dns/privacy
Request domain name, e.g. www.google.com
Request type, e.g. A (which stands for IPv4 record), AAAA (IPv6 record), NS, MX, TXT, etc.
Transport protocol on which the request arrived, i.e. TCP or UDP
Client's AS (autonomous system or ISP), e.g. AS15169
User's geolocation information: i.e. geocode, region ID, city ID, and metro code
Response code sent, e.g. SUCCESS, SERVFAIL, NXDOMAIN, etc.
Whether the request hit our frontend cache
Whether the request hit a cache elsewhere in the system (but not in the frontend)
Absolute arrival time in seconds
Total time taken to process the request end-to-end, in seconds
Name of the Google machine that processed this request, e.g. machine101
Google target IP to which this request was addressed, e.g. one of our anycast IP addresses (no relation to the user's IP)
will not be associated with the Google Account you use for Fiber
This phrase has rather a lot of qualifications which make it trivially easy to circumvent. For example, Google might associate the data with a different account, an internal "ghost" account that you specifically don't use for Google Fiber (but which is identical to it in every way).
I for one would like to see ISPs make an effort to construct systems where even the owner of the ISP doesn't really store any information. For example, they may store differential information about traffic spikes, to help avert DOS attacks. But other than that, they only do what they get paid to do: shuttle bits back and forth across the last mile of internet.
systems where even the owner of the ISP doesn't really store any information
This is basically the default on routers if you don't enable NetFlow. Likewise DHCP and DNS daemons can trivially disable logging. It doesn't cost an ISP money to remove default surveillance; it costs them to add it but the value of the data is worth it.
... information from the use of Google Fiber Internet (such as URLs of websites visited or content of communications) will not be associated with the Google Account you use for Fiber, except with your consent or to meet any applicable law, regulation, legal process or enforceable governmental request.
https://fiber.google.com/legal/privacy.html
They don't really mention what information not associated to a Google account is mined except that it follows the company-wide privacy policy. So they probably mine it. It's probably comparable to Google DNS: https://developers.google.com/speed/public-dns/privacy