Terrorists could use driverless cars to commit attacks
10:43 Share

Hey, it's Adam Grant from WorkLife, a podcast from TED. This episode is brought to you by Freshworks. Freshworks believes that complexity is the enemy of efficiency. So stop wrestling with bloated, expensive service software that takes forever to implement an update, where ROI is someday, not today. You've been overcharged and underserved for way too long. Uncomplicate with Fresh Service for IT and Fresh Desk for customer support.

And with Freshworks AI-assisted service software, you'll work smarter, not harder. Freshworks uncomplicates. Learn more at freshworks.com. As women, we're told to manage our health, track the numbers, bounce back fast. Tia Health is different. They're the only healthcare clinic designed to support women through every life stage, with both in-person and virtual care that's personalized, proactive, and focused on the whole you. At Tia, health isn't the end goal. It's the foundation.

They bring together primary care, gynecology, mental health, menopause, and hormone support, longevity, skin care, and more, all in one place. No more juggling multiple specialists or settling for one-size-fits-most care. Tia's board-certified clinicians build personalized care plans that support not just your body, but your goals, so you can sleep better, feel stronger, and live the life you want. Whether you're navigating perimenopause, managing chronic stress, or

or just want a provider who actually listens, TIA is here to help you live bigger. And they accept most major insurance. Join the thousands of women redefining what healthcare can be. Book now at AskTIA.com. That's AskTIA.com to book today.

Welcome to The World in 10. In an increasingly uncertain world, this is The Times' daily podcast dedicated to global security. Today with me, Alex Stibble and Stuart Willey. It's an alarming warning from the UN that fleets of AI-driven vehicles could be hijacked by malicious actors to carry out attacks. It's brought attention to the potential for these new technologies to be weaponised by extremists.

It's an issue rapidly heading up the agendas of policymakers around the world, given just how vulnerable we all could be to such attacks. With us to consider the risks is Dr William Alcorn, an extremism specialist at Anglia Ruskin University. He's a senior fellow at their International Policing and Public Protection Research Institute.

William, this is a pretty stark warning from the UN Office of Counterterrorism. How worried should we be about the malicious use of artificial intelligence? Yeah, well, I think that it's a good warning. I think we're still in the early days of adoption, especially by kind of terrorist and mine actors. I think that in terms of

threat levels, foreign and hostile state actors are probably the ones to look out for as they have more resources. But I think as the take-up of this technology becomes more mainstream, and we have actually seen terrorist attacks in the past few months that have been assisted with using AI products and also the convergence of different technologies. So in the report, they talk about the autonomous technology

capabilities of vehicles which is not really kind of it is activated at present but might become more of a a realistic threat kind of further down um but we we did see in new orleans at the start of the year a vehicle ramming attack involving an electric vehicle so i think this is something that's that's definitely on the radar of a lot of national security practitioners and policymakers as a potential threat in the future

In the present day, we've seen cars and trucks regularly being used in attacking people in crowds. It's pretty hard to guard against. Would driverless cars make that much worse?

I think that one of the key things in terms of driverless vehicles would be this element of sophistication around kind of swarming. So the use of multiple vehicles kind of chained together and used to attack softer targets seems to be what a lot of terrorist actors have shifted to in the past 20 or so years.

So we can see that as being a kind of plausible possibility, as well as the use of drones as well. The weaponization of drones is another key threat vector, as it were, and being able to coordinate attacks with weaponized drones in public spaces is also something that could potentially be a security threat as well, particularly using kind of pattern-based or facial recognition techniques

technologies and even less sophisticated use of technologies. In the URM report they talk about these examples of someone having about 90 phones and logging onto Google Maps and using Google Maps kind of data and kind of abusing or misusing that in order to divert people to potential targets as well and that kind of pollution of

the training data, as it were, for these AI models. It's a bit more diffuse, but it's potentially something that could lead these models in a more pernicious kind of direction or route.

William, the UN report contains a pretty fearsome list of threats, from biological weapons to swarms of cars smashing into people, potentially using facial recognition to target certain groups of people. Who are the threat actors here? Who are the extremists, the terrorists, who might be interested in this sort of technology?

So mainly we're seeing more established groups and organisations tending to use these. So in terms of

drone warfare. It has been violent non-state actors, not in the UK, but mainly in kind of the Sahel region of Africa. So there's a conglomeration of groups like ISIS, al-Shabaab and other violent non-state actors. I think one of the other kind of aspects in terms of emerging technology is also being use of 3D printed images.

and firearms when it comes to the far right. Many kind of neo-Nazi groups themselves that have hijacked this kind of hobbyist strand of creating kind of freely weapons and used that as a way to

evade detection when it comes to committing potential attacks and atrocities. In my own research where I've seen kind of prescribed organizations talking about using an LLM to ask, what would we need to do in order to create this particular technology? Or what would we need to do in order to conduct guerrilla warfare operations?

We've seen some recent attacks. There was one, there was a stabbing in Finland where the attacker was using even like chat GBT to kind of map out and plan different attack scenarios. So yeah, it's quite...

a mixture between these more formalized groups, but also lone actors. These sort of LLMs, 3D printers, it's tech that most of us can access and is not very sophisticated. Would the threat actors need to boost just how tech savvy they are, or could we see it become a service for hire? The level of sophistication would have to

to increase to such a level that they were able to, especially when it comes to autonomous vehicles, they would have to develop kind of supplementary or kind of alternative technologies and modifications. But I mean, a lot of these technologies have millions of lines of code. It's just, I guess,

I don't want to give them a tip, but it's trying to find the weaknesses within the systems themselves and jailbreaking those systems to then insert new instructions and codes around targeting, around routes, et cetera, and what it can do that can then be used to then maybe coordinate it with other vehicles, etc.

commit one of these atrocities or attacks so it does require somewhat higher level of sophistication but but once that's kind of popularized or known within these circles then then that opens the the possibility in terms of the lowering of the threshold as it were to adoption.

How and how much should countries and their security services and their police be preparing for this sort of threat, do you think? Yeah, I think there needs to be better kind of thought and regulation, not just from government and security services, but from manufacturers as well. There shouldn't be too much of a...

too much of a moral panic around these technologies, but it's also making sure that we're aware of the possibilities that could arise. And I think that's why the UN report was quite helpful, not just to be alarmist about it, but this mode of kind of imagining and thinking about what seems like science fiction stories

kind of future whether it could become a reality I think is is is definitely something for security services and professionals to to think about because I think sometimes the companies might not even be aware of how they can be misused their kind of models or or technologies can be misused so I think it's greater collaboration

I think the way forward is that greater collaboration between the companies and also the law enforcement agencies as well. William, just a possibility for now, but what is the timescale here before this could be a pressing threat? Yeah, I would say that in terms of things like driverless cars, conflation with other kind of technologies and using that for targeted attacks,

it's it's not in the too distant future it's probably in the medium term so i say in the next five to ten years um that could be kind of a possibility in reality especially i know that the uk is slightly being more restrictive in terms of the autonomous kind of capabilities of vehicles but i guess once those become legally regulated that we might see kind of more

proliferation of those technologies within vehicles. So it's not a very kind of long way off kind of goal. I think it's something that we need to think of and that we need to be aware of and designed into these technologies in order to stop some of these more malicious kind of uses and kind of hijacking of

of these technologies. William, thank you. Dr. William Alcorn is an extremism specialist at Anglia Ruskin University. What we've just discussed with William is one of the developing threats facing nations across the world. For the West, another is cyber attacks from China. We analysed this on May the 13th. The episode was called Chinese Cyber Attacks Could Bring the West to a Standstill. Do scroll back and listen to that.

For now, that is it from us. Thank you for taking 10 minutes to stay on top of the world with the help of The Times. See you tomorrow. As women, we're told to manage our health, track the numbers, bounce back fast. Tia Health is different. They're the only healthcare clinic designed to support women through every life stage, with both in-person and virtual care that's personalized, proactive, and focused on the whole you. At Tia, health isn't the end goal. It's the foundation.

They bring together primary care, gynecology, mental health, menopause, and hormone support, longevity, skin care, and more, all in one place. No more juggling multiple specialists or settling for one-size-fits-most care. Tia's board-certified clinicians build personalized care plans that support not just your body, but your goals, so you can sleep better, feel stronger, and live the life you want. Whether you're navigating perimenopause, managing chronic stress, or

or just want a provider who actually listens, TIA is here to help you live bigger. And they accept most major insurance. Join the thousands of women redefining what healthcare can be. Book now at AskTIA.com. That's AskTIA.com to book today.

When you're ordering food for your workplace, are you still using a spreadsheet? Ugh, great. Not with EasyCater, the online platform for ordering and managing food for the workplace. Come on, manually inputting lunch requests in all those little boxes? You don't need that kind of sheet in your life. EasyCater was built for business. It's simple.

Smart and way more powerful than the old sheet you're used to. So drop the sheet and go with Easy Cater, the easy place to get food for your workplace. Order now at easycater.com.