What would happen if botnets were capable of acting together as a collective?
Technology has matured to a point where bots may no longer need to wait for a human to execute attacks and can instead communicate with one another to increase efficiency.
Techday spoke to Fortinet security insights and global threat alliances chief Derek Manky about the concept of swarm intelligence and why Internet of Things (IoT) devices are the path of least resistance.
It’s a strength in numbers concept; by acting together as a cluster and intelligently communicating, computer and robot are able to achieve tasks that they cannot do on their own.
When it comes to examples like ants, they collectively talk to each other and work together to be able to process physical gaps in space, from tree branch to tree branch by forming bridges on their own.
Essentially, multiple agents forge together and conquer to achieve a greater task.
The first idea of Swarm Intelligence was actually coined in 1989.
At that time, it wasn’t being applied to cybersecurity, it was applied to nature but has since been adopted with the rise of robotics and AI.
If we look at traditional botnets and their effect, botnets are monolithic in a sense that they’re somewhat dumb.
They wait for commands from the bot herder (master) in order to execute an attack.
But what if these nodes were able to make autonomous decisions with minimal supervision, use their collective intelligence to solve problems, or simultaneously target multiple vulnerability points in a network using a variety of penetration and exploit techniques?
This is what happens with a swarmbot attack, allowing them to penetrate a network and source intellectual property or data that can be monetised by the attacker.
What are swarmbots and what are the potential impacts that they can have on organisations?
The idea of swarmbots is taking the human attacker out of the loop of botnet attacks, because humans are slow.
If you think of hundreds of thousands of infected machines that can actually talk to each other, they become more agile, quicker and increasingly efficient.
They are able to talk to each other, take action based on shared local intelligence, use swarm intelligence to act on commands without the botnet herder instructing them to do so, and recruit and train new members of the hive.
As a result, as a swarmbot attack that identifies and comprises of more devices will be able to grow exponentially, and thereby widen its ability to simultaneously attack multiple victims
And that’s what we’re starting to see, although in the early stages.
Swarmbots are primarily focused on connected IoT devices as this means there is a significant network which the swarm can move in and try to find the data to infiltrate – it’s a larger attack surface.
The good news is that it’s still at a premature stage right now and we’re only starting to see some indicators now.
In the future, when swarmbots are blended with machine learning models and artificial intelligence, this is going to become a much bigger problem.
The main imminent threat is that as these attacks are starting to happen much faster, they’re more agile and the time to breach (TTB), to launch a successful attack and to take data is becoming quicker.
It’s going from taking hours to seconds.
Already, traditional attacks are taking out critical multimillion-dollar revenue streams; the biggest risk is what will happen when data breaches can happen at a much faster rate.
One of the main problems when tackling this from a cybersecurity perspective is that on the defensive side, putting financial resources into hiring more humans to tackle the problem isn’t efficient anymore.
Almost anyone and everyone is at risk, because swarmbot attacks are somewhat blanketed right now and there is currently very little that can be done to effectively fight off such an attack.
Virtually any organisation that has IoT devices on the network, such as printers and consumer grade routers, is where we are seeing most of the activity and vulnerabilities.
IoT is the path of least resistance currently and where there is significant vulnerability to attacks.
Industries like manufacturing, which have a high adoption of industrial IoT, are more at risk due to these growing connected environments.
There are more security pulls on this type of environment, with more integration happening and more complexities to adhere to.
One of the current inherent issues is that time to breach is rapidly decreasing.
It’s getting to a point where attacks are working at a speed similar to the stock exchange.
Integration, orchestration, and automation is key from a defensive angle.
Automated, intelligent cybersecurity is the direction everything has to go towards.
This will replace the mundane data hacks that a security team deals with and repurpose those people for more complex tasks.
In future, teams can rely on products that can integrate, talk to each other, future operate and having that level of automation and orchestration in their organisation.
The good news is that swarmbots are still in the preliminary stage.
For the first time in 20 years, vendors are at equal arms with attackers because they’re been deploying artificial intelligence and machine learning models for a few years now.
This puts them in a good position to be prepared and stay on equal footing while integrating the cybersecurity changes necessary.