The universality test for moral rules
Kant's categorical imperative asks: can you will that the maxim (the principle underlying your action) be a universal law that everyone follows? If you want to break a promise to escape a difficult situation, the maxim might be 'Break a promise when it suits you.' Could you will a world where everyone follows this maxim? No, because then promises would be meaningless and your own promise-breaking would undermine itself. The test reveals whether your action is genuinely defensible as a principle or merely self-serving.
This is not a rule you follow reluctantly out of fear. It is a test of rationality. A rational, self-aware person should be able to justify their principles to others and to themselves without contradiction. If you cannot will your maxim as a universal law, then your action is morally indefensible, regardless of whether it benefits you or produces good outcomes.
Respect for persons as an end in itself
Kant's second formulation of the imperative states: act so that you treat humanity always as an end, never merely as a means. Using someone to get what you want without regard for their agency and dignity is immoral. This rules out manipulation, deception, and coercion. Even if deceiving someone would produce good outcomes, the deception disrespects their autonomy.
This principle has profound implications. It explains why slavery is wrong (treating persons as mere property), why exploitation is wrong (using people for your benefit), and why democratic consent matters (respecting persons requires honoring their choices). For Kant, morality is fundamentally about respecting the rationality and autonomy of every person, and the categorical imperative is the tool for checking whether your actions pass that test.