In order to answer this question, we first need a definition of the word, "Addicted". Not too long ago, an addictive substance was something that, when taken long enough, produced gross physiological changes in the way the body worked, so that normal operation of the body was impossible without that substance being ingested. And as the substance must, by definition, form a tolerance, higher and higher dosages (up to a point) were needed. This is the definition of "addictive" I'm going to use for this explanation. Addictive is not the same as "habituating". Habituatingsubstances, using this definition, are things you crave, may even come to need, but do not create a gross physiological change in the way your body works (trace neurological and neuro-chemical changes can and do happen but, they're quite minor, and they aren't always substance-related: stroking a pet for instance, can cause such trace effects).
In the cases of alcohol and barbiturates, the addiction, in the sense I describe, is very strong. Stopping these drugs suddenly for extreme addictions usually will require hospitalization, additional medication to treat symptoms of withdrawal and, especially, in the case of barbiturates, may result in death. Lesser addictions like heroin or opiates can also cause withdrawal syndromes, although not as strongly as ethanol or barbiturates, and opiate withdrawal is not fatal (barring the existence of other factors).
The active ingredient in Cannabis Sativa is THC (Delta-1-Tetrahydrocannibinol). THC is active in very low dosages. Therapeutic THC is typically delivered 5mg T.I.D. (three times a day). As addiction in the sense I mean it is a gross process, tiny dosages typically don't generate the large-scale physiological changes a true addiction needs to get revved up (neurological yes; physio no). So most people, scientists and street-users, think of marijuana as non-addictive. A recent study at Columbia University offers potentially contradictory evidence, but it's still only one study and not accepted as universal fact at this time. As such, if you say THC is not clinically addictive, most of the world will agree with you.
Can marijuana be habituating? Absolutely -- but not universally. Just as some people definitely use Marijuana in a manner that can only be described as a habit, some have used marijuana for years but not in a habitual pattern. While the same can be said for alcohol, it seems that alcoholics really do set up a regular pattern of extensive use that I personally don't see nearly as frequently in marijuana users.
In cases of marijuana habituation, I think the causal factors are obscure. With addictive drugs, we can see clear, obvious, repeatable effects in terms of addiction. With marijuana, we see far less predictable results. And why these results are not as predictable is not clear.
The basic fact is that most marijuana uses (maybe all marijuana users) do not display signs of addiction (as defined above).Summary of previous responses:
Many note anecdotal evidence gathered by their own personal use. While this doesn't carry the weight of clinical trials, it also cannot be discounted. Many users reporting here mention what -- based on my definitions above -- would be described as habituation, both mild or strong. Many indicate it's a part of their daily life. None list physiological symptoms caused by withdrawal.
Some mention that people enter rehabilitation, calming marijuana addiction. I would counter by saying that rehabilitation is a cure for a myriad of problems; not just pure addictions, and typically refers to holistic lifestyle change as the solution; not simply breaking the addiction/habituation. So entering rehab in order to address life problems related to marijuana usage is not proof that marijuana is addictive.
Some indicate that they feel THC is addictive when used by those with an "addictive personality". I would counter by saying that, without symptoms of withdrawal syndrome, genetic predisposition for addiction (which is most often used in a context describing a behavioral symptom set rather than a physiological one), does not come into play. No physiological withdrawal symptoms; no addiction.
Some indicate correctly that Cannaibis does not contain nicotine (true), which therefore means it's non-addictive. This syllogism is untrue in that the lack of nicotine only means there's no nicotine addiction in play. This contributor closes by saying THC is non-addictive, which seems to be the general feeling in the scientific community at this time.
It's not physically addicitive in the same way that nicotine is but is psychologically addictive. Modern Marijuana - aka skunk contains much higher levels of Tetra Hydro Cannabinol than the stuff that was around in the 60s/70s.
Anyone smoking this stuff should be very careful.