Dutch companies not ready for AI

Less than a quarter of Dutch companies are ready for artificial intelligence (AI), according to Cisco’s AI readiness index, with only 4% fully prepared to use AI technology.

But according to Marc Engel, partner at international accounting firm Mazars, this is actually not particularly alarming at all. “This technology is developing so fast in such a short time that it is more important for organisations to see through the impact and possibilities now and prepare for it,” he said.

Cisco had an independent research firm survey more than 8,000 companies in 30 countries about their AI readiness, including 159 Dutch organisations. Respondents came from 18 industries and worked at companies with more than 500 employees. The index assessed respondents’ AI readiness across six key pillars: strategy, infrastructure, data, talent, governance and culture.  

Organisations are categorised into “pacesetters” (organisations at the forefront of readiness for AI adoption and integration globally), “chasers” (organisations that are well progressed and above average in AI readiness), “followers” (organisations with momentum towards readiness for AI adoption but below average in preparations), and “laggards” (organisations that are least prepared for AI adoption and integration).  

Only 4% of the Dutch companies surveyed fall into the category of fully prepared pacesetters, and 18% fell into the chasers’ category. This means that less than a quarter of companies in the Netherlands are ready for AI. Some 70% need to be more prepared, and 9% are unprepared. “In comparison, Sweden, which, like the Netherlands, is often in the top five digitisation rankings, is much better positioned,” said Edwin Prinsen, managing director at Cisco, in the blog of Dutch magazine Emerce.  

The biggest shortcoming for Dutch organisations is data availability in the company. After all, this is crucial for the optimal use of AI technologies. However, nine in 10 companies surveyed acknowledge that data in their organisation is too dispersed. Nevertheless, Dutch companies need not immediately worry about the outcomes of the Cisco report, argued Mazars partner Engel.

“Most organisations now have AI high on the agenda – it is worked on, or awareness is high. Remember that it is currently one of the most far-reaching technological and cultural developments. It requires an organisation to be fully ready regarding knowledge, data management, regulation, security and change culture.”

AI strategy

Indeed, Cisco’s index shows that. More than 80% of Dutch companies have developed an AI strategy or are in the process, and more than half state that their organisation is ready to embrace AI soon.

“While it is worrying that almost all Dutch companies surveyed indicate that they are not yet ready for AI, it is reassuring there is awareness,” Prinsen wrote. “Over 90% indicate that the need for AI in their organisation has become more urgent in the past six months and fear that the lack of an AI strategy within a year could have adverse consequences, for example, for IT infrastructure management or cyber security.”

In his blog, Engel wrote that the company is noticing that AI is slowly but surely moving up the agenda of top executives in the Netherlands: “But if we as a country want to maintain our lead in digitisation and prosper economically, it is imperative that there is rapid catch-up in this area.”

Risks of AI

Companies without clear AI policies are at risk, argues Cisco’s chief strategy officer, Liz Centoni. “For now, there is a somewhat tacit agreement for self-regulation to limit the potential dangers of AI, but more nuance is needed on issues such as intellectual property protection,” she said, pointing to using copyrighted information, images or other content by AI.

But bias is also still a significant danger of AI. “The recent EU Data Act already gives consumers more control over their data,” she said. “But as the availability of public data decreases, we expect companies to use synthetic data after 2026 increasingly. Organisations must then watch out for false outcomes and properly train large language models to avoid unintended results.”

Centoni stresses that 2024 will be the year of clear guidelines for responsible internal use of AI. “Until tech companies can credibly demonstrate that they are trustworthy, we can also expect more government regulation,” she said.

Engel expects three-quarters of companies and organisations in the Netherlands to have embraced and implemented AI in the next two to three years. “It is not only an absolute prerequisite from the point of view of competition and customer expectations, but also in order not to fall behind in technological developments,” he said. “It is clear to most Dutch companies that an AI strategy is necessary.”

Training AI talent

Dutch companies still have much catching up to do, not just in terms of strategy, but infrastructure and training AI talent that organisations need to do. Only a quarter of the surveyed companies consider their infrastructure highly scalable for powerful AI workloads. Most respondents say they need more scalability available to operate AI technologies efficiently. “And on top of that, one in three Dutch companies also do not have their data strategy in order, resulting in high data fragmentation,” Cisco’s Prinsen wrote.

The skills, or lack thereof, to deal with AI also represent a new digital divide. Only 4% of Dutch companies have all the skills needed to make AI successful. This puts the country at the bottom of the list of nine European countries surveyed, which also included Germany, the UK, Sweden, Poland, France, Spain, Switzerland and Italy. “And we top the list of companies with no AI skills,” said Prinsen.

While three-quarters of Dutch companies have invested in upskilling their employees, a quarter doubt the availability of sufficient talent to apply AI. Centoni warned of cyber crime risks in this area. “AI-generated disinformation, scams and fraud are fast approaching us. On the dark web, AI is already being used as an attack tool.”

She added that companies and governments would do well to collaborate and invest in risk detection and mitigation, also using AI: fight fire with fire. “New, inclusive AI will have to help fight deepfakes, social media bots, cloned voices and other forms of deception,” said Centoni.

In short, there is considerable work for Dutch organisations to improve their AI readiness.