Cybersecurity starts with the Board of Directors
During the night of 11–12 October, the control units regulating the refrigeration systems in around 700 French supermarkets were remotely tampered with, forcing hundreds of shops to destroy the goods stored in their refrigerated display cabinets. The disaster was not caused by a particularly sophisticated cyber-attack: the target was the control systems of the Danish company Danfoss, one of the leading suppliers of refrigeration solutions to the European retail sector.
Many of the affected control units were at the end of their service life or were protected only by factory default credentials that had never been changed. These devices are connected to the internet for one simple reason: to allow maintenance companies to monitor temperatures and adjust settings without having to send a technician to every shop.
And identifying them does not require any exceptional skills: all you need to do is use Shodan, a sort of Google for devices connected to the internet. The gap between the simplicity of the attack and the scale of the damage reveals a growing contradiction. Whilst attention to security is increasing, companies are introducing a proliferation of procedures and certifications that document the existence of controls without demonstrating their effectiveness: security ends up being better described than it is practised. This is clearly illustrated by the figures from the Cybersecurity & Data Protection Observatory at the Politecnico di Milano: in 2025, around 60 per cent of large Italian companies considered their security budget to be adequate, but only 28 per cent were actually resilient to cyber risk.
The proof lies in the timeline. Critical vulnerabilities in two widely used refrigeration control platforms were made public on 9 August. The attack took place on 11 October: sixty-three days later. On 22 October, Forescout, a company specialising in cybersecurity, identified 12,976 such services exposed on the internet across 99 countries. For the most common type of control unit, 80 per cent of installations were still using a version of the software prior to the one that patched the vulnerabilities. In the weeks that followed, similar attacks affected eight other European countries, including Italia.
It is the gap between compliance and security. The former can demonstrate that a process for rectifying vulnerabilities exists; the latter requires verifying that it actually works, within the timeframe dictated by the real risks the organisation faces. Bridging that gap is not, and cannot be, the sole responsibility of the security manager. In French supermarkets, someone had to decide when to switch off the refrigerated display cabinets, withdraw the products and absorb the losses. In which company could such a decision rest solely with the IT manager?

