As with any security decision, you need a good threat model rather than FUD and giving in to cognitive biases. Whether you consider the threat model in the post to be complete and accurate, just saying "OMG Google could go evil!" is not really a useful way of looking at things.
You need to consider the loss implications if your email is compromised, given various scenarios (the whole account? one message? etc.) and the sorts of information you have in it (journalists working with protected sources vs. griping about your spousal unit malfunctioning). And consider who the likely threat actors would be.
In some cases, hosting on Google would be a spectacularly bad idea. If I were Jacob Applebaum, I wouldn't even think about it, because of concerns about "lawful intercept" and whatnot. But in my case, given the additional security features (two-factor auth, HTTPS everywhere, etc.), and my particular threat model, Gmail does the job just fine.
Comments
As with any security decision, you need a good threat model rather than FUD and giving in to cognitive biases. Whether you consider the threat model in the post to be complete and accurate, just saying "OMG Google could go evil!" is not really a useful way of looking at things.
You need to consider the loss implications if your email is compromised, given various scenarios (the whole account? one message? etc.) and the sorts of information you have in it (journalists working with protected sources vs. griping about your spousal unit malfunctioning). And consider who the likely threat actors would be.
In some cases, hosting on Google would be a spectacularly bad idea. If I were Jacob Applebaum, I wouldn't even think about it, because of concerns about "lawful intercept" and whatnot. But in my case, given the additional security features (two-factor auth, HTTPS everywhere, etc.), and my particular threat model, Gmail does the job just fine.