We went looking for evidence of a pathway from online misogyny to extremism. What we found was considerably more complicated.

In our previous article, When Misogyny Becomes a Radicalisation Pathway, Curious Femme examined emerging research connecting misogyny, the manosphere, particular constructions of masculinity and violent extremism.

Then I kept researching.

And the deeper I went, the more uncomfortable I became with one word.

Pipeline.

It is a wonderfully simple metaphor.

A teenage boy feels lonely or rejected. He searches online for advice. He encounters a masculinity influencer. A recommendation system serves him increasingly misogynistic material. He discovers the manosphere. Misogyny becomes male grievance. Male grievance becomes extremist ideology.

You can almost draw the arrows.

There is just one problem.

That is not what the evidence actually establishes.

There are connections. Some are concerning. Some are supported by substantial research.

But there are also gaps, disagreements and findings that complicate the popular story of boys being pulled down a single, one-way radicalisation pipeline.

So I decided not to force the evidence into the argument.

Instead, this series is going to follow the evidence wherever it takes us.

How this investigation will work

Across this series, Curious Femme will use four editorial evidence labels:

ESTABLISHED
Supported by a substantial or methodologically strong body of evidence.

EMERGING
Credible research points in this direction, but important questions remain unanswered.

CONTESTED
Credible researchers or studies support different interpretations.

NOT ESTABLISHED
A claim may be widely repeated, but the evidence examined for this investigation does not currently justify presenting it as fact.

These are Curious Femme’s classifications of the evidence, not categories assigned by the researchers themselves.

That distinction is particularly important when discussing radicalisation.

Association is not causation.

A statistical predictor is not necessarily a cause.

Two types of content appearing within the same online network does not demonstrate that one causes engagement with the other.

And something being plausible does not make it proven.

With that in mind, what do we actually know?

ESTABLISHED: Gender and misogyny feature within some extremist ideological environments

There is strong evidence that gender is not merely incidental to some extremist ideologies.

Roose and Cook’s analysis of jihadist, far-right and male-supremacist ideologies identifies extreme patriarchal political and social orders, anti-feminism and misogyny across otherwise distinct ideological contexts. Their article argues that male supremacy can also operate as an ideological motivation for violence in its own right (Roose and Cook, 2025).

Research commissioned through the Centre for Research and Evidence on Security Threats, CREST, has separately examined misogyny and hostile beliefs within far-right environments in the UK and Australia. Phelan et al. (2023) examined individual hostile and sexist beliefs, gendered in-group dynamics and broader politicised ideologies, including the role of online channels in transmitting gendered narratives.

Elizabeth Pearson’s research provides another important perspective because it examines gender and masculinity across two very different British extremist milieus: the anti-Islam radical right and the banned Islamist organisation al-Muhajiroun and networks associated with it.

Pearson describes radicalisation as, in part, a masculinity project, examining how particular masculinities can be mobilised through extremist ideology and activism (Pearson, 2024).

That does not mean masculinity causes extremism.

It means ideas about what a man should be can become politically and ideologically significant.

Strength.

Status.

Dominance.

Protection.

Brotherhood.

Control.

Sacrifice.

Different extremist movements can give those ideas very different meanings.

The evidence therefore supports examining gender as part of extremism.

The much more difficult question is how an individual gets there.

EMERGING: Gendered grievance may sometimes become part of a broader ideological explanation

Consider the difference between these statements:

“I can’t get a girlfriend.”

and

“Feminism has destroyed relationships between men and women.”

The first describes an individual experience.

The second offers a political or ideological explanation for it.

Or:

“I feel unsuccessful.”

compared with:

“Modern society has deliberately stripped men of their status.”

Again, something has changed.

A personal difficulty has acquired a social explanation and an identifiable source of blame.

Recent research by Joshua Roose is useful here. Drawing on 80 interviews with Australian men conducted between 2019 and 2024, analysis of anti-feminist and extremist digital networks, and longitudinal national survey data, Roose develops the concept of militant masculinity.

His analysis identifies humiliation, betrayal and perceived male disempowerment as important components within some forms of politicised masculine grievance (Roose, 2025).

This does not establish a universal progression from personal disappointment to extremism.

It does, however, support investigating the point at which personal grievance becomes incorporated into a wider ideological worldview.

Perhaps the important transition is not simply from one website to another.

It may sometimes be the transition from:

This happened to me.

to:

This is happening to men.

and, potentially:

Someone is doing this to us.

What former incels tell us

Some particularly useful evidence comes from researchers who spoke directly to people who had previously identified as incels.

Maryn, Keough, McConnell and Exner-Cortens (2024) interviewed 21 former incels aged between 18 and 38. Participants had first identified as incels between the ages of 15 and 24, with the average age of first identification around 18.

Using reflexive thematic analysis, the researchers identified struggles around masculine expectations, relationships, loneliness and feelings of low personal value among participants’ accounts of entering incel communities.

This is qualitative research.

It tells us something important about the experiences of these 21 participants.

It does not tell us how common those pathways are among young men generally, nor does it establish that loneliness causes somebody to adopt incel ideology.

That limitation matters.

But so does what participants described.

Some had sought help online with problems they were struggling to resolve elsewhere.

Incel communities could then provide explanations, validation and a sense of connection.

That creates an uncomfortable but important question for prevention.

If a harmful community is meeting a genuine social or emotional need, simply condemning the community does not necessarily remove the need that brought somebody there.

Understanding vulnerability is not the same thing as excusing misogyny.

It may be necessary if we want to prevent it taking hold.

People also leave

This is where the pipeline metaphor becomes particularly problematic.

A 2026 peer-reviewed study from the Australian Institute of Criminology examined disengagement from online misogynistic incel communities.

Doherty, Dowling and Dickens (2026) surveyed adult men who had previously engaged with such communities. The survey was open to participants living in Australia, New Zealand, the United Kingdom, the United States and Canada.

After exclusions, the final sample was 43 men.

The researchers found several motivations for disengagement, with disillusionment emerging as the strongest. Disillusionment was also often associated with improved attitudes towards women following disengagement.

Again, this needs qualification.

This was a self-selected sample of former participants responding to an online survey. It cannot tell us what proportion of all incels eventually disengage or establish that leaving a community will necessarily change someone’s attitudes.

But it demonstrates something extremely important.

People can leave.

People can adopt ideas and later reject them.

They can enter communities and become disillusioned with them.

Attitudes can change.

Pipelines only travel in one direction.

People don’t.

Any theory of online radicalisation that explains entry but cannot accommodate disengagement is therefore incomplete.

CONTESTED: What role does “the algorithm” actually play?

This was where I expected the evidence to be straightforward.

It isn’t.

Recommendation systems undoubtedly affect what people encounter online.

But the phrase “the algorithm radicalised him” hides considerable complexity.

There isn’t one algorithm.

Online information environments can include recommendation systems, ranking systems, search results, follower networks, moderation policies, advertising systems, creators optimising material for engagement, communities sharing content and users making their own choices.

Lum and Lazovich (2023), writing for the Knight First Amendment Institute, argue for a system-level approach to algorithmic amplification precisely because even defining “the algorithm” and deciding what counts as amplification requires choosing an appropriate baseline.

The relationship between user and platform is also interactive.

A person watches something.

The platform records behaviour.

The person selects something else.

Creators adapt material to attract engagement.

Communities share content.

Recommendation systems respond to those signals.

People influence platforms.

Platforms influence exposure.

This does not mean recommendation systems are harmless.

It means that moving from “this system affected what somebody saw” to “this system caused somebody to become radicalised” requires much stronger evidence.

For that reason, throughout this series I will generally use the term platform-mediated exposure where that is what the evidence actually demonstrates.

We will examine the algorithm question properly in Part 2.

Then there is looksmaxxing

This is one of the more striking findings in the material we examined.

Looksmaxxing is broadly concerned with improving physical appearance.

At its most mundane, this can involve fitness, skincare, haircare or other appearance-related advice.

In 2025, researchers at the Centre for Emerging Technology and Security, CETaS, published an expert analysis of a large network of looksmaxxing-related content.

Redmond, Small and Hughes (2025) sampled 81,513 videos posted by 45,834 users on a mainstream social-media platform.

They reported that 44 per cent of the looksmaxxing videos they identified also used blackpill hashtags.

Their analysis also identified connections between parts of this network and material glorifying mass shooters, as well as content incorporating neo-Nazi language and symbolism.

Those findings deserve attention.

But we need to be precise about their evidential status.

The CETaS publication is expert analysis/commentary, not a peer-reviewed journal article.

And the findings demonstrate content overlap, network connections and potential exposure routes.

They do not demonstrate that consuming looksmaxxing material causes somebody to adopt blackpill ideology, become a neo-Nazi or commit violence.

So:

Looksmaxxing content overlapped substantially with blackpill-tagged material in this dataset.

Supported.

Looksmaxxing radicalises boys into violent extremism.

Not established.

That distinction is exactly what this series is trying to preserve.

CONTESTED: Is incel violence terrorism?

Even researchers studying incel-related violence disagree about how it should be classified.

Meleagrou-Hitchens and Hughes (2026) examined an original dataset of US federal incel-related cases from 2020 to 2023 alongside prominent attacks.

Their analysis argues that incel-linked violence frequently lacks characteristics they regard as central to terrorism, including coherent ideology, strategic intent and communicative political purpose.

They suggest that grievance-driven misogynistic violence, hate crime or non-ideological mass killing may provide more accurate frameworks for some cases.

Other terrorism scholarship has taken the threat of incel and male-supremacist violence seriously within discussions of terrorism and violent extremism.

The disagreement matters.

Calling an act “terrorism” is not simply a way of saying it was particularly horrific.

Terrorism is an analytical and legal category with implications for policing, security policy, prevention and public understanding.

So Curious Femme will not assume that every misogynistically motivated mass killer is a terrorist.

We will examine the evidence in each context.

NOT ESTABLISHED: Misogyny causes terrorism

After examining the evidence collected so far, I cannot justify writing:

Misogyny causes terrorism.

Nor can I justify:

Watching Andrew Tate causes boys to become violent extremists.

Or:

Social-media algorithms turn ordinary young men into extremists.

Or:

Looksmaxxing leads boys into the far right.

Those claims go beyond the evidence examined for this investigation.

But neither does the evidence justify complacency.

Research has identified misogynistic and patriarchal narratives within different extremist ideological environments.

Researchers have documented forms of politicised male grievance.

Former incels have described pathways involving loneliness, masculine expectations, relationship difficulties and online help-seeking.

CETaS has identified substantial overlap between looksmaxxing and blackpill-tagged material within a large social-media dataset.

And researchers studying violent extremism continue to examine the significance of misogyny, gender and male supremacy.

Something therefore deserves our attention.

We simply need to be precise about what the evidence allows us to call it.

Maybe the pipeline is the wrong metaphor

The popular version looks something like this:

Loneliness → Manosphere → Misogyny → Far Right → Extremism

The research suggests something considerably less tidy.

Personal circumstances can interact with masculine expectations.

Relationships can interact with self-worth.

Grievance can interact with belonging.

Online communities can provide both social connection and ideological explanations.

Platform systems can influence exposure.

People can seek out increasingly extreme material.

Other people can encounter it unintentionally.

Ideas can be reinforced.

They can also be challenged.

And people can disengage.

There may therefore be pathways through this system.

There may be moments when somebody becomes increasingly receptive to extreme explanations.

But there may also be exits.

Different people may move through completely different combinations of these factors.

That is not as satisfying as drawing a straight arrow.

Human behaviour rarely is.

Why this matters for boys

There is another reason I want to get this right.

If we start treating lonely boys, socially awkward boys, angry boys or boys interested in masculinity as potential extremists, we risk creating precisely the alienation we claim to be concerned about.

Loneliness is not extremism.

Romantic rejection is not misogyny.

Masculinity is not pathology.

Vulnerability is not destiny.

The more useful question is what happens when genuine difficulties encounter communities offering harmful explanations.

If somebody says:

“I am lonely.”

and one community dismisses him while another says:

“We understand why. Here is who you should blame.”

then the appeal of the second community deserves serious examination.

The answer cannot be to surrender vulnerable young men to misogynists.

It has to include understanding what healthier alternatives can offer them instead.

So this investigation has changed

I started by asking whether misogyny could become a radicalisation pathway.

I still think that is an important question.

But the evidence has given me a better one:

Under what circumstances can personal grievance become ideological, and what interrupts that process?

That is what the rest of this series will investigate.

Algorithms.

Male loneliness and grievance.

Masculinity.

Looksmaxxing and the blackpill.

Manosphere influencers in classrooms.

The relationship, if any, between domestic abuse and extremist violence.

And the people who actually got out.

Because if we genuinely want to understand radicalisation, studying the people who walked away may ultimately tell us as much as studying those who went further in.

Next in The Misogyny Pipeline

Part 2: The Algorithm Did It, Or Did It?

We hear constantly that social-media algorithms are radicalising young men. But when we examine the research, how much of that claim actually survives?

References

Doherty, L., Dowling, C. and Dickens, M. (2026) ‘Disengagement from online misogynistic incel communities and its implications for attitudes to gendered violence’, Trends & Issues in Crime and Criminal Justice, 727. Australian Institute of Criminology. doi: 10.52922/ti78168.

Lum, K. and Lazovich, T. (2023) The Myth of “The Algorithm”: A System-Level View of Algorithmic Amplification. Knight First Amendment Institute, Columbia University, 13 September 2023.

Maryn, A., Keough, J., McConnell, C. and Exner-Cortens, D. (2024) ‘Identifying pathways to the incel community and where to intervene: A qualitative study with former incels’, Sex Roles, 90, pp. 910–922. doi: 10.1007/s11199-024-01478-x.

Meleagrou-Hitchens, A. and Hughes, S. (2026) ‘(Re) assessing the threat from incel violence: A study of violent incel-related federal cases in America’, Studies in Conflict & Terrorism. Published online 2 February 2026. doi: 10.1080/1057610X.2026.2617209.

Pearson, E. (2024) Extreme Britain: Gender, Masculinity and Radicalization. Oxford: Oxford University Press. doi: 10.1093/oso/9780197772072.001.0001.

Phelan, A., Paterson, J., White, J. and Wallner, C. (2023) Misogyny, Hostile Beliefs and the Transmission of Extremism: A Comparison of the Far-Right in the UK and Australia. Centre for Research and Evidence on Security Threats.

Redmond, J., Small, R. and Hughes, M. (2025) ‘From looksmaxxing to mass shootings: Radicalisation and online misogyny’, CETaS Commentary, 18 August.

Roose, J.M. (2025) ‘Militant masculinity, male grievance, and the politics of democratic decline’, Journal of Gender Studies. Published online 24 October 2025. doi: 10.1080/09589236.2025.2576811.

Roose, J.M. and Cook, J. (2025) ‘Supreme men, subjected women: Gender inequality and violence in jihadist, far right and male supremacist ideologies’, Studies in Conflict & Terrorism, 48(5), pp. 528–556. doi: 10.1080/1057610X.2022.2104681.