The biggest indication this is not malicious is the benefit/risk ratio. If someone at Yahoo decides that talking about Wall st. protests should be silenced, they will, first of all, only silence a small fraction of the traffic - this is the no-single-point-of-failure nature of the Internet. Second, when they get caught (any subject large enough to be meaningful to censor will have enough tech-savvy protractors to detect this) they get hit by both the Streisand effect and the humiliation.
The very fact that the Streisand effect is a well known phenomenon suggests that people running large corporations can't always evaluate benefit/risk ratios very well.
The very fact Streisand is a person and not a company suggests people can't always evaluate benefit/risk ratios very well, regardless of the size and nature of that ratio.
I'd add that the benefit/risk is being evaluated by an individual or a group of individuals, as opposed to the company at large, so what might be insanely terrible for the greater good of the company can also provide a ton of benefit for a specific individual or a group. In that case, the individual or the group would act contrary to the benefit of the company.
It's the classic agency costs argument. I'm actually rather sad that we tend to jump on the company, forgetting that it's individuals that make these things happen.
Of course we can only observe the outcomes where the Streisand effect kicks in - who knows what the risk ratio actually is if one doesn't know the number of outcomes successfully suppressing info...
