On 21 October, dozens of major websites in the Netherlands and Belgium went offline. The cause was two serious vulnerabilities in PHP, the programming language used by around 70 per cent of all websites. PHP is widely used by government bodies, news sites and educational institutions. As a precaution, DigiD.nl, MijnOverheid.nl, the Dutch House of Representatives website, RTL and several hospitals were among those taken offline. In Belgium, all Flemish government websites temporarily went dark.

The National Cyber Security Centre (NCSC) had not yet seen any active exploitation, but urged organisations to install the PHP update as soon as it became available. “Every minute counts,” said deputy director Corine Schipper-Derkse.

AI and cybersecurity are therefore not just matters for the IT department. They affect the whole organisation. In this blog, you’ll learn about cyber incidents and why training helps employees recognise risks and use AI safely.

What is cybersecurity?

Cybersecurity means protecting computers, data and work processes from digital attacks. It’s not just the IT department’s responsibility: employees also play an active role. Organisations protect their computer systems with updates, password policies and access controls. Employees, in turn, need to spot suspicious emails, phone calls and requests. Recognising suspicious activity means employees can take action themselves: don’t click suspicious links, don’t approve requests and hang up on suspicious callers straight away.

Awareness: what are the risks of AI?

What does awareness mean?

Awareness means recognising risks and knowing what to do if you encounter them. AI is playing an increasingly important role in this, because it can make some forms of deception more convincing. It’s not known whether criminals used AI in the incidents described here, but AI can make these criminal activities harder to spot.

How does AI make phishing and fake callers more convincing?

Criminals can use AI to quickly write polished, personalised phishing emails. A phishing email is an email designed to steal your personal data. These emails used to stand out because of their spelling and grammar mistakes, but AI tools can now filter them out. AI can also clone someone’s voice, so a call from a colleague or director may not always be genuine. If you’re unsure about a request or a phone call, check it using another source!

What are the risks of using AI tools?

You also face risks when you use AI yourself. A document or website may contain hidden instructions designed to mislead an AI assistant, for example by getting it to forward information or produce an inaccurate summary. This is called prompt injection: the AI assistant reads the hidden text and follows it as if it were an instruction from you.

There’s also a risk when employees paste customer data or internal documents into a chatbot. If the organisation doesn’t know which AI tools employees are using, it loses track of where the data is going. So only use AI tools approved by your organisation, and don’t paste sensitive data into a chatbot.

Why is training so important?

For AI literacy

Being AI literate means knowing what AI can do, where it can go wrong and how to use it safely. A chatbot can sound confident and still give you incorrect information. So check its answers before you use them.

Think carefully before pasting information into a chatbot, too. Don’t share customer data or confidential documents with an AI tool unless you know your organisation’s rules allow it. The EU AI Act also requires organisations to pay attention to their employees’ AI knowledge. The knowledge needed depends on the role and the situation.

For cybersecurity

Cyber incidents can have serious consequences. They can bring work to a standstill and put data at risk.

On 12 October, a cyberattack hit Albert Heijn’s distribution centres. Shelves were left empty and online orders couldn’t be placed. Etos and Gall & Gall were also affected. Ahold Delhaize filed a police report on 18 October. The company said that, as far as it knew, no sensitive data had been stolen.

According to the reports provided, the incident at Odido started with a phone call. The caller pretended to be an IT colleague and persuaded a customer service employee to enter their login details on a fake website. The data of 6.2 million customer accounts were then stolen.

In France, criminals gained access through outdated software from a supplier and took over supermarket refrigeration systems. Some fridges reached temperatures of up to 50°C. One shopkeeper estimated the damage at around €220,000.

These tactics can be prevented by setting clear rules and providing adequate training.

How does training help you recognise warning signs?

Training helps employees recognise suspicious signals and respond to them correctly. Examples include spotting fake emails, counterfeit login pages and unexpected phone calls.

An AI literacy training teaches employees what AI can do and where the risks lie. They also learn how to use AI tools safely and check answers. That way, they know better when to stop, ask questions or report an incident.

What ground rules can make your team safer straight away?

You can set clear ground rules today. Agree what employees should do when they receive an unexpected request, which AI tools are allowed and how to report an incident.

Sources

• NOS: Government and education websites offline due to vulnerability, risk of hacks (21 October 2026)
• NOS live blog: Websites vulnerable to software flaw, RTL Nieuws, DigiD and MijnOverheid taken offline (21 October 2026)
• NOS: Albert Heijn’s parent company files police report over cyberattack after more than a week (19 October 2026)
• DutchNews: Albert Heijn begins restocking as hack traced to foreign servers (13 October 2026)
• NOS: Cyberattack hits French supermarkets, refrigeration systems set to 50°C (12 October 2026)
• NOS: Odido (12 May 2026)
• Android Planet: Odido hack 2026 (12 February 2026)