For roughly two decades, personality science treated the relationship between dark traits and empathy as close to arithmetic. Machiavellianism, narcissism and psychopathy were associated with empathy deficits. Empathy, in turn, was assumed to restrain harmful behaviour. The implied conclusion was tidy: if someone is warm, socially fluent and emotionally attuned, they are probably not the problem.
In 2021, a team led by Nadja Heym at Nottingham Trent University tested that assumption directly and found it wanting. Using latent profile analysis — a clustering technique that lets groups emerge from response patterns rather than imposing them — on a sample of 991 adults, they identified four profiles. Two were unremarkable: Typicals (34.4%) and Empaths (33.3%). The third was the familiar low-empathy Dark Triad (13%). The fourth, which the researchers named the Dark Empath, comprised 19.3% of the sample and combined elevated dark traits with elevated empathy (Heym et al., 2021).
That group was not a rounding error. It was larger than the classic profile it was supposed to be impossible to belong to.
This article examines what that study did and did not establish. The genuine findings are narrower and more interesting than the coverage suggests. Dark Empaths were markedly less aggressive than the traditional Dark Triad — the gap is large, not marginal — but remained more inclined than Typicals and Empaths towards two specific behaviours: malicious humour and guilt induction. They were the most extraverted of the four groups. They were more agreeable than the Dark Triad but less agreeable than everyone else. And on the vulnerable dark facets — covert narcissism, secondary psychopathy, Machiavellian tactics — they were statistically indistinguishable from the traditional Dark Triad. The antagonistic core survived the presence of empathy. It simply changed how it was expressed.
The article also takes seriously what the popular account gets wrong. Several widely read secondary sources have transposed the prevalence figures. Claims that Dark Empaths judge themselves more harshly than the Dark Triad are not statistically supported by the source data. And the construct itself is contested: Fino et al. (2023) did not recover a clean Dark Empath profile using trait emotional intelligence measures, and Grabovac et al. (2025) raise the possibility that self-reported empathy in high-dark-trait respondents partly reflects self-presentation rather than intact capacity.
For senior leaders and people professionals, the practical implication is not a new label to apply to colleagues. It is a correction to a detection heuristic. Most guidance on difficult personalities trains people to watch for coldness. The evidence suggests that heuristic fails against a substantial share of the population — and that the behaviours which actually distinguish this group are precisely the ones that never reach a formal process.
Warmth is not evidence of safety. Consistency is.