Cyber-threats are an inevitable reality of both our professional and our personal lives, and most of us navigate a range of phishing attempts on a daily basis.

Many of the communications we receive from cyber-criminals are amateurish and easy to discern as a scam which we quickly delete. But the rise of Generative AI means that these communications are becoming increasingly sophisticated.

“At the moment it’s at the higher end – targeted attacks on defence, financial services, critical national infrastructure – that are increasingly difficult to deal with,” says Lee Stephens, Director of Security Services at BT.

“Inevitably democratisation of these toolsets will find its way down the chain. What was originally the preserve of high-end nation-state level hackers can be published as a toolset on a popular developer platform, and then it’s available to anyone who knows how to download it. Generative AI as a hacking toolset is niche at the moment but it will spread.”

Increasingly sophisticated attacks require increasingly sophisticated security protocols. Which make it all the more important that these protocols be frictionless.

Unsurprisingly, security protocols for online access are being scrutinised and adapted to extend corporate and personal protection. Financial services often lead the way in user authentication, and when it comes to protecting our money, most of us willingly submit to two factor or even multi-factor authentication. 

But generally speaking, stringent security protocols in other areas of our lives are jarring, even if the risks are evident.

In recent research conducted by Gartner that looked at the impact of employee behaviour in the context of cyber-security, 93% of users who had demonstrated certain unsecure behaviours were aware that their actions would increase risk to their organisation. This is quite a shocking statistic, demonstrating that when security is a barrier to completing our working tasks quickly and efficiently, it’s human nature to find a faster solution, sometimes irrespective of potential consequences.

Biometrics and single sign on smooth the path, but with a global skills shortage leading to increased workload and increased pressure, frustration at any complex login procedures can result in user workarounds. Even something as seemingly innocuous as emailing documents to a different location – when the on-site printer isn’t working for example – can expose an organisation to attack.

In another recent Gartner report, 69% of employees said that they had bypassed their organisation’s cyber-security guidance in the past 12 months, and 74% said they would be willing to bypass cyber-security guidance if it helped them or their team achieve a business objective.

This breach of our human firewall is exacerbated simply by the fact that human beings are fallible and many of us reuse passwords – the password ‘123456’ has been found 23 million times in data verified by the NCSC. This makes credential stuffing – automatic entry of breached username and password pairs into websites until they find the right account – easy for any would-be hacker with the inclination and the right software.

But the good news is that easily breached passwords, resistance to overbearing security protocols, and user tendencies to find workarounds is driving ongoing innovation in continuous authentication to resolve these challenges. And we can put the hacker’s own toolset to positive use, mimicking and learning their techniques to make fundamental changes in all of our own security strategies.

Predictive or behavioural analytics solutions are now harnessing the power of AI for more incisive, accurate reporting. They offer enhanced monitoring and detection, sending back detailed information to administrators – if anomalies or potential breaches are detected, sophisticated automation enables security policy changes on the fly.

Increasing use of device posture – the collection of security-related data such as OS, browser versioning, encryption and anti-virus status from connected devices – can also help us adapt and enforce our access control policies.

n the quest to smooth user experience without compromising corporate or personal security, there’s a shift towards the practice of continuous authentication – a defining characteristic of zero trust strategy. Instead of a user being logged in or logged out, continuous authentication repeatedly requests verification by assessing ongoing patterns such as user and device location, device posture and other behavioural data, but without requiring the ongoing manual intervention of the user.

Continual computation of an authentication score measures the certainty of the account owner being the one actually using the device and aims to limit both the likelihood and the impact of a breach, without affecting the user’s work experience. There are a number of different emerging technologies which support continuous biometric authentication – by tracking a user’s way of moving, how they hold their device, and how they tap the screen, in addition to facial imagery and voice recognition.

At Adastral Park, home to BT’s Global Research and Development Centre – where we trial and showcase new technology – we’re working on this truly innovative continuous authentication concept. Taking biometric characteristics into account along with physical and data-driven context, we’re currently trialling our capabilities in-house. With approximately 15 patents for this exciting new technology submitted, BT are pioneering the future of continuous authentication so that we can keep protecting ourselves, and our customers from ever-evolving threats.

Security practices and protocols are constantly evolving, and we can’t afford to ever perceive our security strategies as an end state. But by staying abreast of hacker trends, keeping on top of security hygiene, and leveraging the latest in cyber-security innovation, we can minimise the risk to our organisations, quickly remediate any damage, and be ready to repel the latest threats.