Here is my dilettante’s perspective on this matter:
The typical process of evolution involves genetic mutations that randomly alter the genome, occasionally introducing new physiological capabilities that confer an evolutionary advantage to the genome. These changes usually occur on a small scale and can progress over numerous generations. For example, hundreds of millions of years ago, multicellular organisms developed rudimentary neural systems that aided in controlled movement. Subsequently, certain mutations led to the emergence of photosensitive receptors, which added optical sensors to these neural systems. Initially, these receptors and neural networks were minute and quite primitive, but after millions of generations, they evolved into the complex eyes and visual cortex we see today.
The pivotal term I wish to emphasize is "minute". Radiofrequency (RF) necessitates sensors that are many orders of magnitude larger, ideally around half a wavelength. The likelihood of substantial structures spanning meters emerging from random mutations in small, primitive organisms is significantly lower, presumably by many orders of magnitude, than the likelihood of sensors for UV, visible light, and microwave frequencies developing.
To rephrase, this concerns the relationship between wavelength and cell dimensions. The greater the length, the more intricate the mutation that must take place to generate a sensor suitable for that particular wavelength.