Of course not. Google still has that individual's data and can personally identify him. It's just their policy not to use that data when using machine learning and whatnot to get a result out of it.

Also, such "anonymized data" is never really anonymous, until Google starts using something like homomorphic enryption or client-side encrypted databases:

De Montjoye and colleagues examined three months of credit card transactions for 1.1 million people, all of which had been scrubbed of any PII. Still, 90% of the time he managed to identify individuals in the dataset using the date and location of just four of their transactions. By adding knowledge of the price of the transactions, he increased “reidentification” (the academic term for spotting an individual in anonymized data) to 94%. Additionally, women were easier to reidentify than men, and reidentification ability increased with income of the consumer.

https://hbr.org/2015/02/theres-no-such-thing-as-anonymous-da...

The problem is that Google gets that data in the first place and stores it - indefinitely. If someone hacks Google it could also get that data. And what's worse, instead of making it a company priority to "fix" this, it doesn't seem to care very much, even when the solutions could be within reach with a little more R&D:

Lloyd made his pitch, proposing a quantum version of Google’s search engine whereby users could make queries and receive results without Google knowing which questions were asked. The men were intrigued. But after conferring with their business manager the next day, Brin and Page informed Lloyd that his scheme went against their business plan. “They want to know everything about everybody who uses their products and services,” he joked.

http://www.wired.com/2013/10/computers-big-data/