The true distribution of autism in women and men is thought to be about equal. Women and girls with autism are chronically under-diagnosed because diagnostic criteria were originally designed to fit a stereotypical male presentation of autistic traits at the surface level, and the current diagnostic process still hasn't shed that bias.

When women are diagnosed, it is later in life, on average. The "female presentation" of autism (if you can forgive me for not finding a better way to frame that) tends to be less "disruptive" and easier to overlook or explain away or otherwise mask in childhood. Diagnostic methods are already biased and tailored toward children - it is much harder to get diagnosed as an adult. So there's a bunch of pressures against getting diagnosed if you don't fit the mold of the stereotypical (white) male trope of the autistic boy that we all know.

What is known as "autism" is an array of inherent, physical differences in neurology, some of which are still poorly understood, if at all. But we do know enough to know that it's a series of physical dimorphisms that look pretty much the same across the population. Autistic people are born with these differences, and their neurology is defined by them throughout life. Diagnostics always lags behind current science, though. There is no blood test/genetic test that can definitely say "yes, you are autistic". At least nothing practical to deploy at scale. We could theoretically use very expensive imaging, but that obviously would not scale to everyone.

So that's why we see differences not just between genders, but also socioeconomic status, ethnic backgrounds, and other demographic groupings. I haven't even touched on non-binary folks - that's a whole separate ball of wax, and something of a "chicken/egg" situation because autistic people are more likely than the general population to be transgender, agender, or otherwise outside the binary, and vice versa. The point is anyone outside of that stereotypical young white male person born to a middle or upper-class family will have similar challenges with diagnosis, and lower rates of diagnosis in general, just to different degrees, depending on how the traits defining who they are intersect with the diagnostic biases.