Can science predict who’s at risk of radicalising to violent extremism?

The deadly antisemitic shooting at Bondi beach and the attempted bombing at an Invasion Day rally in Perth have put the Albanese government under public scrutiny as Australians demand to know how these acts of terror happened and whether they could have been prevented.

Since September 11, 2001, intelligence agencies worldwide have ramped up their counterterrorism efforts in an attempt to anticipate and stop what many regard as the most feared form of crime.  

Researchers have been instrumental in helping the authorities understand the threats, developing theoretical models that map the pathways leading individuals to extremist violence.

However, even as hundreds of these models inform anti-terrorism strategies today, they still cannot accurately predict who is at greatest risk of radicalisation.

Associate Professor Emily Corner, a terrorism researcher at The Australian National University (ANU), explains why.

“Current risk assessment tools are retrospective and rely on anecdotal evidence of known cases. They are based on one terrorist or a group of them, and each single model is different with little consistency,” she says.

“A lot of them focus almost exclusively on terrorists who are Muslim, which means they cannot be applied to other contexts. As a result, prevention efforts have followed the same narrow lens by focusing on Muslim communities, leading to a lot of issues that have impacted social cohesion across the world.” 

Terrorism is recognised by experts as a global phenomenon, rather than an issue exclusive to any single religion or community. In recent years, it has become increasingly ambiguous –with data revealing nearly two-thirds of recent attacks couldn’t be tied to a specific group or ideology.

Also, the relative youth of terrorism studies as a field has made progress difficult.

“Most sciences have been around for hundreds of years. Terrorism research is about 50 years old and only really gained momentum after 9/11,” Corner says.

“That’s when national security practitioners introduced the concept of ‘radicalisation’. Researchers then had to work backwards to define and study it because no one really knew what it was.”

All of this, along with the low incidence of violent extremism events, have reinforced the belief that radicalisation is unpredictable.

But is this assumption substantiated?

Predicting the unpredictable

From marketing to medicine, science is today capable of predicting human behaviour with surprising accuracy.

“Supermarkets are laid out in a specific way because they can predict how humans shop,” says Corner.

“Similarly, medical research can predict rare mental health diseases such as schizophrenia.”

According to the ANU researcher, the same predictive logic used in these fields could transform how experts assess radicalisation.

However, most protocols countering radicalisation and terrorism work very differently.

They rely on so-called ‘antecedents’, factors that preceded a person’s shift toward extremist beliefs or behaviour in the past and that are often mistakenly treated as risk indicators by practitioners.

Social isolation is one commonly cited example. In 2015, 18-year-old Australian Jake Bilardi died in Iraq carrying out a suicide attack on behalf of the Islamic State (IS). His father later described him as a shy and lonely teenager who struggled to fit in with his peers.

Corner cautions that the presence of one or more risk factors in an individual is not conclusive evidence of future violence. Just because these factors influenced the choices of Bilardi and others doesn’t mean someone else with similar traits will inevitably follow the same path.

“About 99 per cent of people considered radicalised will never enact an act of violence,” she says. “Risk factors aren’t causative.”

“There are also protective factors, which reduce the likelihood of radicalisation happening. These, too, must be considered by practitioners.”

To account for the true complexity of the human psyche, Corner emphasises the need to study risk and protective factors in real time across the general population, rather than simply looking back in time.  

“This can be done through formulation assessments, commonly used in medical research, which look at identified risk factors in the context of each individual and their circumstances to determine if they’re truly relevant,” she says.

“Unfortunately, formulation needs to be done by trained psychologists, and the prevention of terrorism sits largely within national security departments, where psychologists are a minority.”

A model with predictive power

Poised to address deficiencies in her field, Corner, together with other terrorism and psychology researchers from ANU, have developed a novel framework that has shown promising signs of predictive power.   

“Most existing models of radicalisation are unidirectional, with a starting point and an endpoint in radicalisation,” she says.

“They’re easy to interpret, but they only work with past cases that experts have examined retrospectively.

“Our model works as a loop – focusing on the ongoing, reciprocal interactions of three core elements of a person’s life: our cognitive processes, our social context and our behaviour.

“In other words, how we make sense of the world, the people we engage with and what we do.”

Corner believes the model could serve as a much-needed foundation to help support more sophisticated policy to counter radicalisation and terrorism around the world.

Is Australia doing enough?

While predicting radicalisation is not possible yet, governments can’t afford to wait. So is Australia doing enough to prevent violent extremism?

“I don’t think any country is doing enough,” Corner says. 

“Governments, police and intelligence services are actively trying to prevent terrorism, but terrorists are creative and they always find new ways to circumvent surveillance systems.”

Australia’s terrorism threat level remains ‘probable’, meaning that there is a greater than fifty per cent chance of an onshore attack in the next twelve months.

Corner argues that the only way to eradicate radicalisation is to adopt a public health approach – a goal, she admits is practically utopian.  

“We’d need to treat terrorism as a public health problem like we do with disease management,” she says.  

“Such an approach would require inoculating entire populations and focusing on people who are at risk of developing the ‘disease’.

“Seeing real change would take generations and require engaging people at a young age, ensuring we’re protecting and meeting their social and health needs.

“This is a pipe dream. There is no appetite to put that much investment in such actions because priorities change as new governments take office.”

For now, much of the preventive work falls to community-based programs.

Initiatives such as Living Safe Together give family, friends and community members avenues to report threats.  

Corner says these are essential in an environment where lone offenders are thriving. According to a March 2025 report, 93 per cent of fatal terrorist attacks in the West over the preceding five years were carried out by lone actors.

“Lone actors pose a challenge because they’re not physically engaging with a group. Intelligence services rarely lack awareness of organised extremist groups, but individuals who operate mostly online are much harder to detect,” she says.

“Western governments rely on community reporting tools.”

If you’re concerned that someone you know is at risk of involvement in violent extremism, call the Step Together helpline 1800 875 204.