![](https://dl6pgk4f88hky.cloudfront.net/2021/06/4-scaled.jpg)
Cyber threats have been a reality almost since computers were invented, but they’re evolving. People carry their data with them in their pocket. The “Internet of Things” means that apparently inanimate objects can connect and communicate, and some companies are allowing their employees to use their own devices.
You might imagine this would cause the nature of threats to change dramatically. In fact, according to Stuart Aston, Microsoft’s chief security adviser, a lot of the dangers are well established. The good news is that a lot of the techniques for fighting back are well established, too; the bad news is that not everyone is adopting them.
First, the new stuff. The Internet of Things isn’t actually all that big a deal, believes Aston. “Any time we provide a device with some sort of connectivity, we should think about the security of that device – what it can do, what it’s connected to, what it knows and what it doesn’t know,” he says. “Then we should think about how it’s appropriate to secure that in a risk-management way.”
In other words, as other contributors have said in this supplement, it’s worth assessing what the risk actually is and how dangerous the information a device might hold can become. Smart light bulbs that allow themselves to be turned on and off remotely are undoubtedly connected, Aston points out, but if they’re hacked then the consequences are hardly disastrous. “Then the bad guy knows whether my lights are on or not. Is that a big threat? Do I care?”
The answer may well be yes, if the light bulb’s connection can be piggybacked to gain access to banking or other confidential details, but it might be completely harmless. The key is applying security technology intelligently, as distinct from the mainstream idea of aiming simply for ubiquitous cover. In business, in particular, it has to be a cold, detached decision.
“If I had a diamond in my house and I didn’t have locks on the doors, the bad guys could help themselves,” says Aston. “But if the cost of the locks and all the other security was more than the value of the diamond, I have to start asking how badly I want the diamond.”
Up in the cloud
The other idea that could provoke some disquiet is that of putting everything into the cloud. This is actually a strength, Aston believes. “We spend something like a billion dollars per year on cyber security at Microsoft, which is a big investment,” he says. “I doubt many commercial organisations can afford to spend that much resource on making their own services as secure.”
There are follow-on benefits from taking this approach, he believes. You start off with one customer whose installation of, say, Microsoft Office 365 is under attack in the cloud. Microsoft works out what’s going on and how to prevent it, and protects not only that customer but all those that use Office 365. “We use highly trained people to look at security events and react to them accordingly,” Aston says.
Cloud, when it’s done right, also helps companies in sensitive environments. “People know where the information is stored, how it is stored, and then you as the consumer of that cloud service can make a judgement about your risk – it does become a choice, because cloud is about choice. Many consumers are observing that and saying they’ll put a portion of their data into the cloud for a well-evidenced resource.”
The main documented threats still come from the traditional avenues, Aston explains. These are examples of disguised malware, which might look like an email or a video codec or something else, which the end user – the criminal hopes – will install on their system, believing it to be innocent. It then starts doing something else on the network: “It could install software designed to steal banking information. It could be a Trojan designed to download other malicious programs. It could be a root kit on a PC. It could be there for ransomware. It could be there to spy.”
Ransomware is a relatively new development. Frequently delivered by a Trojan, which is just the mechanism, it establishes itself on a system and cuts off access to data unless a ransom is paid. Trojans have declined as a menace over the past year: Microsoft research shows that around 3.5 per cent of computers with reported malware have Trojans, as distinct from 5 per cent a year ago, but longer-term data suggest this risk is fluctuating rather than dying down.
Browser modification is another recent development. Here, a web browser is changed to make it show an unwanted advertisement (which sounds harmless, if annoying); or to record keystrokes (less harmless if you’re typing in sensitive passwords); or any number of other things. At the same time, criminals are moving towards “social engineering”, observing how someone behaves on social media and when emailing – their signature, their general manner and so forth. They can mimic this behaviour, so that recipients of any messages become convinced they’re communicating with a friend or colleague. This friend or colleague then turns out to be tranded at an airport and needs a money transfer, or something similar.
Rise of the robots
The social engineering phenomenon points towards humans being part of the problem. This is true to some extent. Easy-to-guess passwords and default security settings left on phones and other devices are a widespread security risk. (Remember the phone hacking scandal a few years ago – in which phones weren’t actually hacked. The criminals simply guessed that the owners wouldn’t have changed their voicemail passwords.) Today, people are talking about security more than they used to and are increasingly in the habit of reporting incidents, says Aston, which has to be a positive thing.
“In the Seventies and Eighties there was a spate of [fraudulent] doubleglazing sales that were made over the phone,” he adds. “The callers asked for credit card details, and people were giving their details over the phone. So the person is also part of the attack.” Such scams are less common today, but can still occur.
However, automation is a major part of the equation. “The reality is that you’re looking at millions and billions of security events every day,” Aston says. “You can’t have someone going through all the code by hand and saying, ‘That looks a bit fishy.’ You have to generate machine-based algorithms to work out what’s out there. It turns out those algorithms are about 100 times more efficient than the humans, anyway.” There is a great deal of machine-tomachine learning happening, which helps sharpen the systems’ responses. So, logically, does this mean that the criminals can also use machine-tomachine learning? Aston suggests that the cost would be prohibitive, although the theoretical possibility exists. And with state-sponsored cyber crime now well established, it might be asked just how finite hackers’ financial resources could actually be in some cases.
But fighting the malware remains relatively straightforward, Aston says. Leaving the actual cure for the viruses and other malware to the giants such as Microsoft, there are simple practical steps that individuals can take, and that companies can train their employees to implement. “It’s quite common for identity to be used as part of an attack, and it’s a simple thing to protect against, or at least be aware of,” Aston says. “You can do a number of things. You can use multiple factors of authentication; you can use an authenticator; you can use smartcards or other mechanism.”
Staying up to date
The other thing to do is to ensure that all software is up to date. “Many customers don’t have up-to-date software on their PCs, so the bad guy gets a free pass,” says Aston. “That’s not just from Microsoft’s point of view – every piece of software needs to be kept up to date.” This is normally achieved through automatic patching, although many customers find these facilities annoying and switch them off. This magnifies the problem not only because the security hole is unplugged but because there are known patches to address specific vulnerabilities. The cyber criminal knows there will be systems in the field with that specific problem, so they know where to find the weaknesses.
“It’s like hygiene,” Aston says. “We go to the toilet; we wash our hands. We go to a hospital ward; we wash our hands or give them a spray. The chances of our getting sick are much reduced. It’s the same with updates and generic software.
“What we see for the UK is something like one in eight computers report an encounter with malware over a six-month period. Also in the last quarter, 3.5 per cent had Trojans, down from 4 per cent last year. Browser hijackers were at about 5 per cent for the fourth quarter of 2015.”
The figures might sound low but, given the sheer amount of computers in the wild, they’re not. It’s arguably reassuring that the emergence of cloud, the Internet of Things and other innovations is unlikely to damage security, but it remains the case that the basics around identity are often ignored. Overcome this and, while it’s unrealistic to expect it to go away, the threat will at least be mitigated.