While Kant’s ethical arguments are of the form “thou shall not X because if everyone did X the following bad things would happen”, Objectivist principles are of the form “man’s nature requires doing Y, if he does not-Y he dies, thus man should do Y if he wants to live”.
Actually, you’re pretty off on this one. It’s a common misconception that the categorical imperative restricts actions because “bad things will happen.” That would be inherently consequentialist and, therefore, anti-Kantian.
What the CI requires is for the individual to formulate a maxim. Something like “When it is beneficial to do so, I shall lie.” Kant says this can’t be rationally willed, because not everyone can lie. Lying requires others to tell the truth at least sometimes, otherwise lying won’t work.
Now, that’s the basis of the first formulation of the CI. The second is a little less arcane. Instead of being about universalizing of wills and maxims, it just states “Always act so that you never treat another solely as a means.” So I can’t lie because I will treat another person solely as a means.
One fundamental difference you might want to give thought to is in Epistemology, which is the “bridge” between Metaphysics and Ethics. While Kant proposes ethical commandments based on the imagined consequences of those commandments, no more than a thought experiment, Objectivism derives ethical principles from the nature of man - objective facts of reality.
Again, Kantianism rejects consequences as justification for ethics. The bridge between metaphysics and ethics lies in Kant’s views of the nature of man, as well. He sees the rational will as having infinite value. The person, he says, is “beyond price.” Actions are just or unjust based on the rational will behind them (the maxim). Thus, the maxim must be capable of always being adopted (being consistent with the way moral statements are made, ie “Do X.”) and must not violate the sanctity of another’s will.
As we learned about Kant, one hypothetical situation popped into my head.
That thought experiment, famously enough, was proposed by several of Kant’s contemporaries. Essentially, they proposed that you are confronted with someone running from a murderer, who then hides in your house (with your consent, of course). The murderer comes knocking and asks where his victim is. It would seem that the CI would demand that we tell him the truth, or at the very least not lie.
Kant gives a rather stupid answer that is, overall, pretty inconsistent with his previous statements (By the time this was proposed formally to him, he was getting on in years and fairly senile). But I think the answer lies in Kant’s right to punish. He says that when a criminal commits a crime against us, we have a duty to punish them in proportion to their crime. We can do this primarily because we are treating them according to their own will (hurt people!).
We have a right to ask, “Why do you ask me this?” If he lies (and we are fairly certain he is lying), we can freely lie to him back as punishment. If he tells the truth, we can simply refuse to tell him where his victim is. Furthermore, we can freely lie to him because of his own violation of the CI.