Can AI Predict Your Sexual Orientation?

If sexual orientation is determined by a complex interaction between nature and nurture, one might predict that this interplay of forces should lead to recognizable differences in brain anatomy and physiology. Although many studies have used brain measurements of hetero- and homosexual individuals to discern potential characteristics, it has been difficult to draw valid and robust conclusions due to the small sample sizes of most studies and the variability of experimental designs. Newer resting-state functional connectivity (RSFC) fMRI methods are easier and cheaper to conduct, have a much better signal-to-noise ratio, and allow for a substantially larger suitable population of participants.

Using this RSFC methodology, a recent study investigated whether sexual orientation can be reliably predicted, based solely on a brief five-minute brain scan, using machine learning programming and predictive pattern classification. Machine learning is a type of artificial intelligence (AI) that gives computers the ability to learn without explicitly being programmed.

The subjects were male and female homo- and heterosexuals. The results demonstrated that the newer RSFC methodology and machine learning AI programming were able to correctly classify 82.9 percent of homosexual and 100 percent of heterosexual individuals. Five brain regions were identified that accurately and consistently predicted heterosexual or homosexual individuals. Four of the five regions were in the frontal lobes, two in the left frontal cortex, one in the right precentral gyrus, and one in the right orbitofrontal cortex.

One outlier was in the right precuneus cortex, which lies in the parietal cortex in the back half of the brain. The precuneus is connected to brain regions that process visual and pheromonal stimuli as well as sexual arousal. Thus, it seems plausible that the precuneus, in combination with the prefrontal cortex, is involved in processing aspects of sexual orientation. [If you would like to understand more about the brain, please read “The Brain: What Everyone Needs to Know.”]

This study demonstrated for the first time that a simple five-minute resting-state scan of the human brain can, with high accuracy, successfully predict sexual orientation in healthy participants. Sexual orientation is a highly complex human trait. While the differences in brain measurements between homo- and heterosexual participants allowed for a relatively accurate prediction of sexual orientation, the results of this study do not address the question of what caused these differences, such as the effects of nature or nurture.

Previous studies in male subjects have speculated that the connectivity patterns observed are the result of genetic factors. In contrast, no significant genetic linkage has been discovered with regard to female sexual orientation. Overall, genetic differences do not appear to be involved in sexual orientation. No genome-wide association studies, comprising genetic data from almost 500,000 individuals, has thus far lead to a meaningful prediction of an individual’s sexual orientation.

In conclusion, using a state-of-the-art machine learning AI approach, this recent study demonstrated how well sexual orientation can be predicted by neurobiological measures of brain functional connectivity.