Hasty generalization?

At what point does a generalization become fallacious and hasty? If i say people who smoke get cancer and the data shows that %90 of smokers get cancer, is that still hasty generalization? Or %60? Would that be hasty generalization? And what about %50 and %40? 

At which point can i I make the statement that "people who smoke get cancer"

A hasty generalization , like most informal fallacies, exist on a continuum. This means that are "degrees" of fallaciousness, of which are arguable. One also must take into account the context in which the argument is being made. For example, if someone were to say, colloquially, "people who smoke get cancer," this would be a fair and reasonable assessment of data as long as it is understood as a generalization. Of course, not everyone who smokes get cancer, so this isn't technically or scientifically accurate. Given this, within a more structured debate context or especially a scientific context, it would never be acceptable to say "people who smoke get cancer." One should be as precise as possible in these contexts and choose words carefully. For example,

90% of people who smoke more than a pack of cigarettes per day develop lung cancer after twenty years of continued smoking.*

So in summary, fallaciousness for this particular fallacy is dependent on intention, degree, and context. An argument would have to be made for any claim as to how fallacious it is or is not.

* This is an example, the data is most likely inaccurate. Scientifically, I would also want to know if this was one study that showed this? Who are the "people" - demographics? It this from a meta-analysis, etc.

The issue is in the way your statement is worded. Regardless of whether smoking increases the risk of cancer by 40% or 90%, all you can say is, smoking increases risk of cancer. You cannot say "smokers get cancer" because some don't. To understand the argument, people have to understand probability.