AI hackers in a ‘huge’ threat landscape - new book troubleshoots security and privacy for 6G wireless technology
Friday, 28 July, 2023
Listen to the (opens in a new window)podcast
(opens in a new window)Madhusanka Liyanage is Ad Astra Fellow and Director of Graduate Research at the School of Computer Science at University College Dublin. His new book (opens in a new window)6G Security and Privacy - a Comprehensive Guide troubleshoots the sixth-generation wireless technology slated for roll-out in 2030.
While some of us are still figuring out 5G - well, fifth generation wireless technology is still in the deployment phase - others, like Dr Madhusanka Liyanage, are already looking a decade ahead to its next iteration, 6G.
But if 5G is the fastest, most robust technology the world has ever seen - with quicker downloads, lower lag and a significant impact on how we live, work and play - why will we need 6G?
“The world is evolving, we need different types of services and the current mobile network might not be able to support these technologies. The metaverse - a virtual reality space in which users can interact with a computer-generated environment and with other users - is a good example of this."
“Initially I thought the metaverse was something really hard to realise and very far away. But when I see the recent development and investment from companies like Facebook, Amazon and Accenture, I think we will be experiencing at least the first version of the metaverse very soon.”
The metaverse promises immersive experiences beyond our wildest imaginings in the material world, like flight and space travel. Artificial intelligence will play a much bigger role in future wireless networks that enable complex services like the metaverse.
“AI is used on some of the 5G applications but in 6G AI will be integrated into every part of the mobile network, including the devices, the network itself and also all the applications which are running on top of the safety network. So 6G is AI-based networking that can be considered an autonomous network run with minimum or no human interaction. 6G is not executed yet but theoretically we have a lot of building blocks that can be used to support new services.”
Liyanage outlines three potential scenarios where 6G privacy and security will need to be monitored closely.
“One is that an attacker can use AI to attack the 6G network, because AI-assisted attacks or AI-based attacks are now emerging, where the AI itself and not humans are doing the attack. AI cyber attacks are going to be a big thing in future,” he adds.
Secondly, the AI system integral to 6G is considered “a new attack base” and can itself be disabled through data poaching or data model poisoning attacks.
“And the third one is, I would say, the benefit: we can use AI to protect the system. We have to use AI-based security mechanisms to fight against the AI-assisted attack.”
Liyanage’s book offers solutions to these so-called ‘zero day’ or hypothetical attacks.
“AI can help us to identify these kinds of AI attacks before they happen. Because in 6G, we are talking about very critical services like autonomous cars; if somebody attacks those it can have fatal effects. So that means that we cannot wait until the attack happens before we do the mitigation.”
Current 5G security issues will be “more intense” in the 6G world. For instance, the security functions on many Internet of Things (IoT) devices - such as phones, laptops, virtual assistant technologies, smart TVs, smart fridges - are often already insufficient.
“Millions of these devices are being made and many IoT device manufacturers do not have the expertise to install correct security functions in their systems. That's why people can attack these IoT devices very easily. In 6G there will be billions of IoT devices, so the attacks initiating from the IoT devices will be amplified. That's another challenge.”
Conspiracy theories sprang up around 5G, erroneously conflating its deployment with the spread of Covid-19. Similarly unfounded rumours are likely to emerge about 6G, but Liyanage expects these to be mostly AI-related.
“People who might not trust the AI system. ‘How can you ensure this AI system is following the policies and regulations defined by humans and in an ethical manner?’”
The solution, he says, is to educate people about AI ethics and Explainable AI “so they understand how the AI system works”.
The potential benefits of the metaverse include improving online learning, education and business opportunities and negating physical distance. Meanwhile 6G might finally enable long-promised advancements like telesurgery, which uses wireless networking and robotic technology to connect surgeons with patients who are distantly located from one another.
But the 6G threat landscape is “huge”, which is why Liyanage and many others are already working - almost ten years ahead of its deployment - on making 6G safe.
“We are not going to deploy these things blindly. We know what is going to happen and we have our tools to fight against them,” he says. “Now we are struggling to protect the physical world. On top of that we will be protecting another virtual world; not one world but multiple worlds on top of that. I think most of the security issues we see in the physical world will be directly applicable in the metaverse. For example, somebody stealing your identity or property; digital properties can be stolen in the metaverse. There can be other technology-related issues like denial-of-service attacks, when somebody cuts the data stream so you don't have access to your avatar or control over it,” he says, of your virtual persona. “It will be a big challenge to secure the metaverse but I won't say impossible.”
Does it worry him?
“No, actually, I'm excited. As a researcher, we are always excited when new things happen. There are a lot of things we can do to mitigate risks - and that excites me.”
Listen to the (opens in a new window)podcast
This article was originally published on 29 September 2022.