Dark empath: what the research actually found

A dark empath is someone who scores high on the dark triad traits, narcissism, Machiavellianism and psychopathy, while also scoring high on empathy.

That combination is the whole point. The usual assumption is that dark traits come with an empathy deficit. This profile does not work that way.

Almost everything written about it calls the dark empath “the most dangerous personality type”. The study that produced the term found something different, and if you are trying to work out who you are dealing with, the difference matters.

Where the term comes from

One paper, in 2021. The Dark Empath: Characterising dark traits in the presence of empathy, by Heym and colleagues at Nottingham Trent University, published in Personality and Individual Differences.

They assessed 991 people for narcissism, Machiavellianism, psychopathy, and both kinds of empathy, then used latent profile analysis to see what natural groupings fell out. Four did:

Group	Share of sample	Profile
Typicals	331	Low dark traits, average empathy
Empaths	357	Low dark traits, high empathy
Traditional dark triad	128	High dark traits, low empathy
Dark empath	175	High dark traits, high empathy

So the dark empath is real as a statistical profile, found in roughly one in six of that sample.

It is not a diagnosis. It is not in the DSM-5 or the ICD-11, and no clinician can tell you someone is one. It is a research description that escaped into popular use.

What the study actually found

This is the part that gets left out.

Both dark groups were more aggressive and less agreeable than everyone else. That much matches expectations.

But comparing the two dark groups directly, the researchers found the dark empath had higher extraversion, higher agreeableness, and lower aggression than the traditional dark triad group. The dark triad group also showed poorer wellbeing.

Their own summary: the dark empath “is less aggressive and shows better wellbeing” than the traditional dark triad, “but partially maintains an antagonistic core”.

Read that carefully, because it cuts both ways.

Against the headlines: on the measure most people care about, aggression, dark empaths scored lower than the classic dark triad. The “most dangerous personality type” framing is not what the research says.

Against reassurance: they did not differ from the dark triad on the narcissism facets, and the antagonistic core is still there. Less aggressive is not the same as safe.

So why does it hurt so much?

Because empathy is not one thing, and the split is what makes this profile distinctive.

Cognitive empathy is knowing what you feel. Reading the room, tracking your mood, predicting your reaction.

Affective empathy is feeling it with you, and being moved to do something about it.

The same research group had already mapped this split. An earlier paper on empathy deficits across the dark triad examined which empathic subtypes are tied to dark traits.

It also looked at which ones carry indirect relational aggression: the sideways kind that works through reputation and exclusion rather than confrontation.

A dark empath tends to have the first. What varies is whether the second translates into anything you would recognise as care.

That produces a specific experience: being understood with unusual accuracy by someone who does not act on that understanding in your favour. They know precisely which comment will land, and they make it anyway.

With an ordinary narcissist you often feel unseen. Here you feel seen, and used with the accuracy that seeing provides. People frequently describe that as more disorienting, not less.

What are the signs of a dark empath?

• Uncanny accuracy about your feelings, including things you have not said out loud.
• Warmth that arrives when something is needed and cools once it is not.
• Jokes aimed at the exact place it hurts, followed by “you know I’m only teasing”.
• Guilt rather than force. The pressure is emotional, and hard to name afterwards.
• A reputation for kindness that makes your account of it sound implausible.
• Extraversion and charm, which the study found in this group specifically, so this is rarely the withdrawn type.
• Difficulty with criticism, met with hurt or withdrawal rather than open anger.

The reputational point is the practically important one. This profile is sociable and well-liked, so your account of it sounds unlikely.

The isolation that follows is the same as with a covert narcissist. Nobody else sees it, so you conclude it is you.

Is a dark empath a narcissist?

Not necessarily. Narcissism is one of the three dark traits measured, and this profile involves elevation across the group of them alongside high empathy.

The clearer contrast is with the traditional dark triad. Same antagonistic core, but with the emotional radar switched on and, per the study, less overt aggression.

	Narcissist	Dark empath	Traditional dark triad
Reads you accurately	Sometimes	Yes, precisely	Often not
Overt aggression	Varies	Lower	Higher
Social manner	Varies	Extraverted, warm	Often abrasive
Acts on your feelings in your favour	No	No	No
A diagnosis?	NPD exists	No	No

Can a dark empath be a good person?

This is the most-asked question about the term, and the honest answer is that the question is slightly wrong.

The study measured traits, not moral worth, and a profile drawn from a questionnaire does not determine how a person behaves toward you over years.

The useful question is not what they are. It is what they do, repeatedly, when it costs them something. Traits describe tendencies. Conduct is the evidence.

And you do not need to resolve their character to act on your own experience. If the pattern is that your vulnerabilities keep being used, that is sufficient information, whatever label fits.

Which part of this actually predicts harm?

The 2021 study’s own phrase is the one to hold onto: the dark empath “partially maintains an antagonistic core”. There is separate evidence on what that core does.

A preregistered meta-analysis of 118 studies broke narcissism into components and tested each against aggression.

Interpersonal antagonism was positively associated with every index of aggression tested. The outgoing, socially confident component was associated with less aggression, not more.

That maps onto this profile uncomfortably well. Dark empaths scored higher on extraversion and agreeableness than the traditional dark triad, and lower on aggression, and none of that touches the antagonism underneath.

So the sociability is not protective. It is just not the risky part. What you are assessing is how much contempt sits under the warmth, which is visible in conduct rather than in manner.

What do you do about it?

The label changes nothing about the response, and the response is the same as for any accurate reader who uses what they read.

Stop supplying material. The advantage here is informational. Someone who knows exactly which memory lands hardest can only use what you have given them, which is what the grey rock method is for.

Watch what happens after you disclose something. A person with affective empathy softens. A person with only the cognitive half files it. The difference shows within a few weeks.

Do not test them with vulnerability. People try this, hoping to be proved wrong. It hands over more material and answers nothing you could not learn by watching ordinary conduct.

Judge by cost. Anyone can be kind when it is free. What matters is what they do when being decent costs them something they wanted.

Expect to be understood and not spared. This is the specific disorientation of the profile, and it is worth naming, because being read accurately feels like intimacy right up until it is used.

Do not wait for the label to resolve. You will not settle whether someone is a dark empath, a covert narcissist or simply unkind, and none of those changes what a repeated pattern is telling you.

A caution about the label

“Dark empath” spread through social media faster than the research behind it, and most of what circulates has drifted a long way from a single 2021 paper.

You will see it presented as the rarest and most dangerous type, as a diagnosis, and as something identifiable from a quiz. None of that is supported.

It is worth holding lightly for a practical reason too. Labelling someone is not the same as deciding what to do, and the search for the exact right term is often a way of postponing the second thing.

Is a dark empath the same as an empath in an abusive relationship?

No, and the confusion between these two is worth clearing up. They are close to opposites.

Search results and social media use “empath and narcissist” to describe a pairing. A highly sensitive person ends up with someone who exploits that sensitivity. That is a real dynamic, and it describes the target.

A dark empath is not the target. It is the other role. Someone with an antagonistic core who also reads emotion accurately, and who uses the reading rather than being harmed by it.

The same trait, applied in opposite directions.

	The sensitive partner	The dark empath
Reads emotion	Accurately	Accurately
Uses it to	Anticipate and soothe	Locate leverage
Cost falls on	Themselves	The other person
Outcome	Exhaustion, self-erasure	Influence maintained

Two things are worth saying about the “empath” label itself. It is not a clinical category, and there is no assessment that identifies one. What people usually describe is high emotional sensitivity plus a habit of managing other people’s states. That is a learned pattern rather than a fixed identity.

The distinction matters practically. An identity cannot be changed. A habit can.

The reason so many people arrive at this material describing themselves as an empath is that the habit was trained early. Usually in a house where reading the adult was the safest available strategy.

If you recognise yourself in the sensitive-partner column, the useful reading is self-abandonment, which covers the pattern and how it reverses.

Common questions

Is dark empath a real diagnosis? No. It is a research-derived profile from a 2021 latent profile analysis, not a clinical category.

Are dark empaths the most dangerous personality type? That claim is not supported by the study it usually cites. On aggression, dark empaths scored lower than the traditional dark triad group.

Is a dark empath just a covert narcissist? They overlap and can be hard to distinguish from outside. The practical experience is similar enough that the covert narcissist article will likely be more useful if you are trying to describe what is happening to you.

Can you spot one with a test? No. The online quizzes have no validation behind them.

What do I do if I am close to one? The same as with any pattern of manipulation: watch conduct rather than explanation, stop supplying reactions, and set boundaries that describe your own behaviour rather than demanding theirs.

What to take from this

If someone in your life reads you accurately and uses it against you, the label is the least important part of that sentence.

The research is genuinely interesting, and it is worth knowing that it says something more careful than the headlines built on it. But you do not need to place someone on a personality taxonomy before you are allowed to protect yourself.

Where to go next. The covert narcissist is the closest neighbour to this profile, and narcissistic supply explains what the warmth is being spent on.

Communal narcissism covers the version built on visible kindness, and can a narcissist love? covers what the accurate reading of you is actually for.

How to set boundaries with a narcissist is where the practical work starts.

If you want structured support, the 8 week narcissistic abuse recovery program is built for the aftermath of exactly this kind of relationship.