Data, AI, and Cyber Security
Tom Cooksey
News
8 min read

In September 2026, emails and webform messages from 12,000 residents of a major housing association were accessed following a cyber-attack on its website [1]. Residents were advised to contact the housing association by phone or email while services were affected.
It wasn’t the first incident. In 2021, its repairs provider, Plentific, suffered a data breach which resulted in residents receiving phishing emails asking for cryptocurrency payments for repairs [2].
The question for block managers is no longer whether cyber-attacks affect their sector, but whether their software suppliers are prepared when one comes.
Big names aren’t immune
These breaches pale in comparison to Capita’s, in which 6.6 million people’s personal data was stolen [3], and the 2025 JLR attack, which halted production for five weeks and caused an estimated £1.9bn impact across the UK economy [4]. It just goes to show that large organisations with significant technology resources are not immune to cyber-attacks.
How do we stop it? The British leasehold management sector is comparatively small and relies on several small software providers.
That may mean we’re less targeted, and the quality of UK in-house developed software is generally high. But the worrying new trend is products built entirely on AI: vibe-coded software which might appear convincing, but what goes on under the hood is terrifying.
AI has also lowered the barrier to entry for hackers. Nefarious actors can now instruct swarms of agents to probe software for vulnerabilities cheaply and quickly. The focus used to be purely on organisations where the potential reward justified the high effort. That equation is changing rapidly, and now everyone is fair game.
From my perspective, AI has increased the fixed costs of developing software properly. At RiskBase, our spend on infrastructure and security, plus external specialists and contractors monitoring our systems, represents a substantial and growing investment. These aren’t optional extras; they’re part of the cost of operating a serious software business.
The growing risk for block managers
A few decades ago, compliance in block management was closer to financial record-keeping. Now there are countless electrical, gas and particularly fire-safety requirements to manage. Hackers may not care about compliance itself, but they can steal personal data to sell to fraudsters or disrupt systems and demand a ransom to restore them.
Legislation, including the new residential evacuation plan regulations [5], also means block managers hold more personal information about residents, including sensitive details about disabilities and evacuation requirements. Health information, including details of disabilities, is special category data under UK GDPR and needs additional protection [6]. The reputational risk of a breach is huge.
The questionnaire illusion
Most of our clients aren’t security specialists, and why should they be? Customers care about our product, how it works and how it interfaces with existing software. For many, there’s a general assumption around security.
Others often send fairly one-dimensional information security questionnaires. We complete them and tick boxes around ISO 27001 certification and CREST accreditation. A supplier can have ISO 27001 certification and still suffer a serious breach; certification tells you something about how security is managed, not that every system is secure.
There are good questions in those questionnaires. The problem is that you’ve given out the exam questions. Security is often best tested in conversation, putting someone on the spot rather than letting them go away to revise. It should also be an ongoing exercise, not a one-time event.
Backups are essential in the case of data loss, but they won’t stop a ransomware attack. In the Capita attack, nearly a terabyte of data was copied before ransomware was deployed [3]. Backups didn’t undo the data theft or make the ransom go away.
Performing a backup is only one part of the solution. How often is it tested to make sure it works? There have been numerous examples where companies have failed to do this; even Pixar wasn’t immune when it was making Toy Story 2 [7].
Five questions every block manager should ask
You can’t control every organisation in your supply chain, but it’s not hard to have a basic conversation with a software supplier and see whether they inspire confidence. Get someone technical on a call and ask:
When was their last penetration test, what did it look for, what did it find, and what have they done about it?
What security tools do they use? How do they check their code and the infrastructure it runs on? What happens when a tool finds a problem?
How are vulnerabilities fixed? How are issues identified and critical updates prioritised?
What data do they hold? What information can the supplier access, and who can see it? Do they hold sensitive personal data about your clients or residents?
Can you get independent advice? If you don’t have the technical knowledge to assess the answers, bring in a specialist.
Finally, make sure you have cyber insurance with a sensible excess. Check your policy covers incident response, recovery costs and business interruption. The supplier should also have cyber insurance and specialist technology errors and omissions (E&O) or professional indemnity (PI) cover.
I’ve been writing software and managing engineering teams for almost 25 years. I see what goes into keeping software safe and secure, and I’m not convinced that our industry knows what to look for. I think that needs to change.