On 4 December 2025, teissTalk host Thom Langford was joined by Ed Tucker, Director - Cyber Security Practice, Telefornica Tech; Neil King, IT Security Professional, Canon; and Christine Bejerasco, CISO, WithSecure Intelligence.
In 2026, the companies that do well will be the ones that see cybersecurity as a strategic pillar for the whole business, not simply an IT cost centre. A lot of the problems (AI threat vectors, quantum risk and synthetic identity) need coordination at the business level, as well as board engagement and a shift in culture. What executives need to do is make the CISO a strategic business partner. Gut instinct and reporting when anyone feels something isn’t right remains key, and applying current protection capabilities to future threats can improve resilience. As attacks are getting multichannel, so should defence. The article doesn’t mention identity as the next frontier – especially in the browser. Meanwhile, incident response must speed up, as currently some vendors’ logs are delayed up to 24 hours. Responsibility for what AI agents do is going to be another problem area. Who will own their identity if they misbehave?
Ransomware will remain the biggest, most organised and destructive threat. If you have to rebuild an entire infrastructure network, it’ll take months, which impacts business continuity. Therefore, it’s key to establish what your minimum viable business is. Ransomware’s impact has increased so steeply that today no businesses can ignore it. It can even wipe out the better part of a nation’s GDP.
Nowadays simple software, such as BitLocker is being used for encryption in ransomware attacks, which can cause a lot of headache to security experts, as well as improve “customer experience”: unlocking, although tricky, is guaranteed – unlike in the case of proprietary criminal encryption. Vendors can also be sometimes the problem as some of the leading ones still can’t offer capabilities that would flag up two consecutive log-ins within minutes that are geographically unviable to perform. If technology blocked processes that social engineering tries to take advantage of, people on help desks simply couldn’t perform actions that criminals want them to. Moreover, executives should always go through the same security protocols that anyone else in the corporate hierarchy must to thwart criminals attempts at social engineering.
Nation-sponsored cyberattacks are another major challenge. The number of North Koreans posing as a new remote worker to infiltrate corporate systems is on the rise – as baking tax and identity that will pass most checks is becoming easier by the day. Recruiting more locally can mitigate these risks, though but recruiting from a global talent pool has its appeal too. Hopefully, in 2026 LLMs and AI can be better harnessed to look after a company’s own data, enabling businesses to create their own threat intelligence, which will lead to a better understanding of one’s own corporate threat landscape. We’ll probably see more sanity around AI deployments, with governance playing an increasingly central role. On the geopolitical threat side, we are most likely to see more intolerance for nation-sponsored attackers than before. Meanwhile, threat intelligence may get to a quality level in 2026, where it can be leveraged for the automated patching of production systems.
Read about rickrolling here.
© 2025, Lyonsdown Limited. teiss® is a registered trademark of Lyonsdown Ltd. VAT registration number: 830519543