A recent op-ed in the New York Times advocated for a law requiring bystanders to help people in emergencies. It says that such laws are common in Europe, but only a few US states have them. The value of such laws would lie mostly in the social message they send, since the penalties are usually weak and they're probably not strictly enforced. The article notes that people are more likely to judge an action to be immoral if it is illegal. Perhaps these laws (in places that have them) could be used to argue against omission bias in general.

With the exception of a few jurisdictions, the “no duty” rule remains largely the same as it was famously described by William L. Prosser, the dean of American tort law: “The expert swimmer, with a boat and a rope at hand, who sees another drowning before his eyes, is not required to do anything at all about it, but may sit on the dock, smoke his cigarette, and watch the man drown.”

[...]

A sensible statute might read like this: “Any person who knows that another is in imminent danger, or has sustained serious physical harm, and who fails to render reasonable assistance shall be fined up to $5,000, imprisoned for up to three months, or both.” Civil liability could also be established, as in other countries.

This "sensible statute" sounds as though it could describe most of the population, if it were to be given a utilitarian interpretation.

I am not an attorney, but I do know that trying to help a person who has a medical problem can put the helper in legal jeopardy. For example, if somebody who is far from a hospital or other sources of help looses a leg in an accident, the person will die without a tourniquet. If I touch the person in an attempt to help, maybe that person will get an infection from me or charge me with battery. Legally, I would have to get the person's permission before trying to help unless the person is unconscious. If I misinterpret the situation and apply a tourniquet when one is not needed, then I could be sued for killing the person's leg, even if the person gave me permission to help.

I was in a situation not long ago in which a homeless person was at risk of death (starvation or exposure) due to his mental illness and the harsh conditions where he was staying (basically he had periods of paranoia in which he refused donations of food). I tried to get the police to send him for a 3-day psych evaluation, but they would not touch him unless he was unconscious. I could not touch him without committing a crime (battery). The homeless person eventually hitch-hiked out of town, and I don't know his status. The question is, what, exactly, is "imminent danger," and how is an untrained lay person supposed to determine what this is? It is very difficult to determine when a person might starve to death or die of heat stroke.

Every US state has a Good Samaritan law, but as you said, it's hard to be sure whether you would be protected by it (or maybe you live somewhere without such a law). Where I live, you're protected from liability if you attempt to aid "an injured, ill, or emotionally distraught person who reasonably appears to the person rendering the aid to be in immediate need of emergency aid in order to avoid serious harm or death." Was the homeless man in immediate need? Maybe not. And the law is quite conservative about involuntary treatment of people with mental illnesses, perhaps to their detriment.

Psychology is not a science and the concept of mental illness is neither logically coherent or scientific. It's legitimately astounding that people take the concept of mental illness seriously and that psychiatrists have the legal power and perceived authority that they do but even if you believe that mental illnesses are matter of factly real the more important question is what's wrong with being mentally ill. If you're a ('philosophical') hedonist you can't consistently regard mental illness as inherently harmful or bad unless you're using that term interchangeably with suffering. The actual felt emotional state of distress alone is not what psychiatrists have in mind when they diagnose people with various disorders (the concept of psychiatric diagnosis is illegitimate in itself at least partly because the mental states of other people are inferred, not directly / inter-subjectively observable, but I don't want to focus on why psychiatry is matter of factly an illegitimate science) and a hedonistic understanding of 'mental illness' can't justify setting apart the 'mentally ill' as a distinct group of people because you can't draw a non-arbitrary line between people who feel whatever specific 'form' of suffering intensely and very often and those who experience it less intensely and less often - the nature of pain doesn't change when you increase the amount of it. If 'mental illness' isn't distress itself then it's something that causes distress at specific moments in time which is still a fundamentally different position than the concept of mental illness that exists among psychiatrists today. Some delusions and hallucinations have a neutral or positive impact on the felt emotional states of some 'schizophrenics' and if I'm not mistaken the drugs used to treat psychosis work by lowering dopamine and a common (if not inevitable) side effect is depression and a reduced sensitivity to pleasurable stimuli. Euphoria is also sometimes diagnosed as pathological (mania).

If you're a preference consequentialist then you have to reject the idea that mental illness is of objective inherent negative value and you have to consider someone's mental illness to be bad only because they consider it to be bad (which isn't to deny that preference utilitarianism can justify paternalism - it can if a preference utilitarian believes that frustrating one desire a person has is necessary in order to fulfill other desires they have but attempting to minimize mental illness because it can cause something harmful or bad and 'treating' it for it's own sake are fundamentally different positions).

I think that paternalism is sometimes justified (by 'paternalism' I mean violating someone's autonomy for their own benefit ) but you have to factor into consideration the humiliation, frustration, self-doubt and general emotional distress that it causes ( people may vary in terms of how much humiliation - and in the case of psychiatric incarceration /forced treatment/ unwanted diagnosis' etc. at least ; self-doubt - they feel as a result of involuntary psychiatric treatment or paternalism in general but the felt frustration of your desire to do what you want to do with your body and your life or to live where you want to - ie. in the comfort of your own home, is necessarily distressing). Psychiatrists tend not to do this since the 'mental illnesses' they are preoccupied with are only loosely or vaguely, if at all, related to moment to moment emotional states (among other reasons).

Being detained by force in a psychiatric ward or hospital and placed under the control of people who tend to be condescending and lacking in any real compassion or empathy for their patients will not improve the felt quality of life, immediately or in the long-term, for arguably the majority of people who are detained against their will in psychiatric wards.

This in itself isn't an argument against psychiatric incarceration but I find it strange that you can legally detain someone for attempting suicide or not 'taking care of themselves' and force psychiatric treatment on to them but you can't force people to accept life saving blood transfusions or certain cancer treatment or detain them for choosing not to do so for various religious, philosophical or personal reasons.

I'm not opposed to good Samaritan laws in principle but, as with all laws, you should factor into consideration the harm government coercion will cause those who break them or would be inclined to do so. I left out some details partly because I don't have time but I have more to say on this issue.