It’s true; the future of cyber security is AI.
It’s advancing daily, and eventually, all antivirus and general security protection will be provided by AI-enabled analysis. But we’re some way off this yet, and despite advances, AI is only one element in building our security defences against future cyber threats. It is important to take a risk and business-centric approach to gathering and examining threat intelligence and making informed decisions on this at Board level.
Risks to an organisation can emerge despite having made significant investment in security controls. We can become complacent once that initial investment has been made and forget that buying a tool is the beginning, and not the end, of the journey. SIEM products provide the best example of this. So, we have a big tick on the spreadsheet next to ‘security’, but does it really mean that the organisation’s defence is any better?
The effectiveness of the product is conditional upon the organisation’s depth of expertise in being able to tune the solution to the specific and properly defined protective security monitoring objectives. It’s rare that organisations have the necessary in-house skills to be able to do this and if they do, you can be sure they will be ‘headhunted’ before too much longer. This complacency, of just investing in software and thinking that it will meet all our requirements ‘out of the box’ rather than understanding potential threats and how the product can help us to manage these, is likely to be leaving your organisation open to attack.
Sector-specific security threats
Risk is coming from all around and your defence may not be as strong as it can be because – it hasn’t been managed correctly; it hasn’t been tailored to your environment; strategic decisions haven’t been made, or you bought a product and you don’t completely understand how to get the best out of it.
With much confusion still existing around cyber security, and an industry that is driven by selling the latest security products, leaning solely on technology to address threats to your organisation is an easy mistake to make. To an extent, everybody should have the same base level of security to protect themselves against the most prevalent attacks. This is why the Government came up with schemes like Cyber Essentials to help with cyber security training. But once you have this baseline, you then need to understand where your threats are coming from and tailor your defence accordingly as well as identify any gaps you might have.
It may be that you are over-investing in technology because you based your understanding of your threats on what the salesman told you. Rather than every organisation deploying the same level of technical security, the best outcome is achieved when security is tailored to your environment. The best way to get started on this is to consider a sector specific approach to risk management.
Different sectors, such as defence, banking, health, criminal justice, retail and manufacturing, will be targeted by different threat actors. For example, if your organisation is in the medical research sector, the organised attacker would be more interested in gaining access to your data, than to sting you with a ransomware attack. Put yourself in the shoes of the attacker to determine what information within your organisation they will be most interested in.
It may be that a nation-state has an interest in the research data meaning that your threat defences should be geared towards preventing access to this using a tailored set of controls that are over and above what would be referred to as ‘baseline’. Whereas organisations with business or mission-critical services, such as scheduling surgery in hospitals to save people’s lives, may be more at risk of ransomware as attackers will be counting on them to pay the ransom to minimise the impact.
Building a clearer picture of the threat
To build an effective defence, you need to know what the risk to the organisation is at any given time. Draw on as much information as possible on the threat to your environment. Information can be gained from a range of sources, such as stakeholders in the business, industry sector or by other interested bodies such as the National Cyber Security Centre which provides sector-specific threat information.
It’s not IT’s (sole) responsibility to define information security policy or a cyber security strategy. Quality information needs to make its way firmly into the Boardroom to drive proper protection of critical assets and the risks that these assets face. Effective presentation and interpretation of that information provides the intelligence for someone to make a strategic decision which then informs what kind of defence you need, and where investment in endpoint security technology should be applied. Following this, we can analyse the pros and cons of different anti-virus, SIEM or AI products.
When you take this approach, you have a far better handle on security rather than being driven by whatever technology is the flavour of the month. Decisions are ‘top-down’ rather than ‘bottom up’ which also supports the requirement for improved Board level ownership and accountability. If you know which types of attack are more likely and by whom – a hacker, a competitor or a nation-state – you’re in a far better position to build a defence against them.
AI – the best defence or the advancing enemy?
There is a dark mirror image in the security industry which is less mature now but advancing all the time. As quickly as AI is being developed into a security solution, it will also be used as a powerful threat against us.
AI will make malware more intelligent and increase the likelihood of successful exploits. If malware can learn from every occasion that it is repelled, it won’t be long until it learns how to circumvent our static defences and finds a different path to infiltrate the organisation. This poses a major challenge to malware prevention. In other words, AI is going to increase the likelihood of a successful attack, so we need to be ready for this, and signature-based malware defence products are really going to struggle against this scenario.
Imagine if the AI-enabled malware breaches the corporate defence and sits on the network, watching and learning every bit of information going in and out of everyone’s inbox. Not only would important corporate financial data be at risk – social and personal information would be too. For example, it could learn the names of your children, what your interests are and what you are planning to do at the weekend. Imagine a phishing attack tailored to your personal information, referencing your daughter’s hockey match at the weekend, containing a picture of her scoring a goal that you had already e-mailed from your phone to your work account to use as your desktop background.
The malware poses as a friend or colleague, presents you with that great picture and tells you to click on the link in the e-mail to see lots more images from the match. How much more likely is this to be successful than a random e-mail from an unknown source simply saying: ‘you will like this’?
Hacking as a Service
Hacking as a Service, while not a new concept, is predicted to become much more prevalent in the future. Whether for personal or commercial interest, wanting to hack another’s account is being fuelled by the social media explosion. Where previously you could visit a hacking site to buy a hacking tool, now you can buy Hacking as a Service from malware engineers and people that specialise in breaking into profiles.
Hacking services are becoming much more professional with money-back guarantees and whilst of course, hacking remains illegal, these services continue to be a direction of travel for people with that motivation. What’s more, the ability to buy Hacking as a Service rather than a tool to use yourself, takes the person one step back from the crime – and therefore makes it more appealing.
Time for the Board to catch up
The days where simply ‘ticking along’ and relying on the IT service management team with your cyber security practices are over. The ongoing day-to-day level of good security practice must continue, but more needs to be done at a strategic level to determine where and how future threats will be targeted. There is a huge degree of catching up required from the Board to improve its understanding of cyber security.
This is not because Boards don’t want to get involved, it’s simply because security has not traditionally been a core business governance activity in the way that say finance, sales or marketing have been. Boards are simply not confident in this area and need to be properly supported. Significant GDPR fines like the £183 million one BA received earlier this year should certainly help to see cyber security noticed at Board Level. This should result in better linkage between the business and its security, and hopefully, appropriate investment too.