If a client is paying with insurance the T needs to give a diagnosis. Aside from that it is part of the service a client is paying for. If they don't use insurance then it's a bit different and a diagnosis isn't required. If the client wants a diagnosis, the T will give one, but they don't have to.

At an intake appointment with an outpatient T, clients are generally asked why they are seeking therapy in the first place. The questions that follow are based on that answer. So if a client answers that they are depressed, the T will ask questions about symptoms associated with depression (as outlined in the DSM). If parents bring a child to therapy to treat ADHD symptoms, the T will ask the parents and child questions based on symptoms for ADHD (as well as questionnaires).

Ts will also typically ask about lifestyle habits and figure them into a diagnosis: sleep, eating, substance use, social supports, family, friends, school, work, etc. In my experience, unless a client reports symptoms indicitive of something more complex than depression or anxiety, Ts won't come to such a conclusion after only one session. Ts also won't base a diagnosis off of observation alone - they can make a note of it and keep observing, but that's all.

So, I think in many instances, clients have some amount of input into a diagnosis that is more complex than a mood, anxiety, or adjustment disorder. And it should serve a purpose- whether it's to simply get insurance coverage or to ensure the proper treatment. Ts don't, or I should say are not advised to, assign a diagnosis if it is of no benefit to their client.