See past discussion at https://forum.inaturalist.org/t/survey-about-criteria-for-wild-captive-observations/27007/10. Some broad take-aways:
There seems to be rough consensus that if a specific organism was introduced to a location by humans, that’s a good indication that it may be captive/not wild. However, this should not automatically apply to its offspring. This fits well with existing iNat official guidance.
Being fed by humans isn’t a good criteria, but receiving care or cultivation from humans, or being dependent on humans for survival, is.
Criteria about restriction to a particular area are only useful if that area is relatively small (or smaller than the organism’s natural territory). For example, just because Kruger National Park in South Africa has a fence around it, doesn’t mean the animals should be considered captive/not wild.