The current consensus appears to be that Autism Spectrum Disorder (ASD) has a multifactorial and ultimately largely unknown aetiology. See e.g. Sauer et al. 2021.
This question will resolve YES if a satisfactory causal explanation of ASD is found by midnight, Dec 31, 2026.
The explanation would not necessarily apply to all cases that are currently classified as ASD based on current diagnosis best practices, but to at least half of them, based on my understanding of the medical consensus as summarized by the UK NHS website.
For reference, the NHS website currently states that "Nobody knows what causes autism, or if it has a cause". The question will resolve when the NHS website will instead report a statement such as "It has recently been found that ASD type-H is caused by the combination of a genetic predisposition and exposure to cadmium in utero" with "type-H" accounting for at least half of the individuals diagnosed with ASD.
Note that the above example is entirely fictional and used for illustration purposes only. Also note that if ASD stops being considered a disease, this does not impact resolution (non-diseases can still have causes). If people will no longer be diagnosed as ASD but .new sub-classes or an entirely new classification scheme is introduced, this question will still refer to what was considered ASD on Aug, 4 2023. The finding of an effective treatment, per se, does not count towards resolution.
If the NHS will no longer provide such a statement about ASD, will cease to exist or to operate an official website, or will be convincingly shown to be supporting a minority view with respect to scientific consensus, I will rely on a suitable equivalent.
Would it something like this count? "ASD type X (about 30% of cases) is caused by [X]. ASD type Y (about 25% of cases) is caused by [Y]." That is, multiple distinct causes are identified which combine to include at least half of the cases. I'm guessing this would count in the spirit of the question but I wanted to make sure.