When importing a raw image to PhotoShop, the Camera Raw dialog lets you adjust the white balance as well as other settings. Now, adjusting the white balance at this stage makes perfect sense, is it will always be a conversion of some sort, and I can see how it can be useful to adjust the exposure when doing this conversion to keep as much information from the image as possible, but why are the defaults set to adjust blacks, brightness and contrast? Surely the default should be to trust that the exposure is as intended?
The blacks, brightness, and contrast aren't a matter of exposure. They're adjustments to the tone curves. A neutral curve represents the data in a flat way. This is true to the pixel sensor data from the camera, but looks lifeless and dull to the human eye.
So, a more interesting curve is set to produce a higher black level and more contrast. This reduces some of the data in the image from the pure neutral form, but it looks a lot better. Selecting the curves that look the way you want them to is one of the essential functions of RAW conversion (in addition to demosaicing and sharpening).
Many cameras have settings for different "picture styles" in the camera. That's how you select the different curves you want for in-camera conversion from the sensor to JPEG.
With film, the different tone curves are inherent to the type of film, and are affected somewhat by development choices. Being able to make this decision after the fact — or even in-the-fact but with flexibility shot-to-shot — is one reason people like to work with digital.
I don't know ACR, but it appears that the defaults are selected to produce something that looks more appealing than a flat, neutral RAW file as a starting point. This makes sense, as otherwise you'd have one million newbies complaining that the product sucks and produces awful-looking images without a lot of work. It's my understanding that you can tune the ACR defaults to your liking.
Keep in mind that a raw file isn't an image; it's just data from the sensor that isn't even pixels until it's been interpreted. Everything in the image you see is a conversion. Even the decision as to what constitutes a pixel is a conversion. ACR or any raw processor, including the one in the camera, constructs an image by combining data from adjacent R, G and B receptors in very complex ways to derive pixels.
Here's the thing: there are lots of different ways to process the raw data to come up with an image. ACR is one, and the default image it produces is based on a guesstimate of what most people will find pleasing.
As mattdm says, the ACR defaults include Adobe's guess at pleasing tonal curves for the R, G and B channels and the composite of all three. On the Camera Calibration screen of ACR, you can pick any of the camera manufacturer's curve presets, or you can "roll your own" using the Basic, Tone Curve and HSL screens.
Just as an averaging light meter guesses "18% gray" no matter what the actual image looks like, ACR makes its guess based on a somewhat arbitrary "average" image. If they tend not to work for your images, you can save new defaults that fit your work better.