AI Act 2026

One thing is for sure: there will be no way around AI in the future. And as has been the case for any other fundamental technological development in the past, it will present companies with both opportunities and risks. There are few enterprises left today that do not (yet) use AI: be it through their employees (secretly) relying on ChatGPT or the IT department introducing new Microsoft functions like Copilot or implementing special AI applications as software solutions.

What is the AI Act?

Most companies already know that the use of AI entails regulatory, ethical, and social responsibilities vis-à-vis a wide variety of stakeholders. Up to now, however, many of the related efforts took the form of voluntary company commitments. From 2026, the EU will take things a step further with its AI Act. After the act has entered into force, companies will have comprehensive compliance duties – and that’s not limited to AI providers but applies to all enterprises using AI.

The final text of the world's first regulation on artificial intelligence is currently being finalized, and the act will enter into force in 2026. Its key points: It specifically bans AI use in some areas, among them the categorization of people based on sensitive features such as religious belief. This is the highest of overall four risk levels assigned to the various AI applications. In addition to this “unacceptable risk,” meaning that AI must not be used, there are “high risk,” “limited risk,” and “minimal risk” with varying stipulations regarding AI use in companies.

Is GDPR history Repeating Itself?

As of now, managers have less than two years left to strategically prepare for the arrival of the AI Act and avoid unnecessary costs and work as a result of ignoring the anticipated measures.

“With the AI Act, I am reminded of the run-up to the GDPR: everybody knew that the regulation would come. Everybody knew that strategic and practical efforts were required to prepare for that day. But most companies still buried their heads in the sand. They simply hoped that things would somehow sort themselves out,” says Barbara Stöttinger, Dean of the WU Executive Academy, who advises executives across all sectors and industries to start looking at the fine print of the AI Act as soon as possible.

After all, no company will remain unaffected, just as has been the case in past major technological waves, such as the widespread availability of the internet in the 1990s or social media in about 2006.

Your Own AI Strategy: an Investment Worth Making

The EU made the decision to deal with this topic in its own way and introduce more demanding ethical standards than the rest of the world – which could very well benefit European business players. It is hoped that B2B and B2C customers will regard ethical and transparent demeanor in the use of AI as a valuable competitive edge.

The following 6 modules can form the foundation of a detailed AI strategy business leaders can rely on to prepare their organization for the AI Act.

1. Get to Know AI

“Understanding the basic functions of AI applications is a must for any manager today, not only to understand the ramifications and (potential) outcomes of AI use but also to keep up with and speak the language of staff members in the IT department. Business leaders need to understand both the consequences of their decisions (in a world abounding with AI) and what it would mean if competitors used AI while their own company doesn’t. They need to know as much as they can,” Stöttinger says. At the WU Executive Academy, the topic of AI has been covered as a major component of the digital economy in many courses, where it is discussed, strategically analyzed, and, in the best cases, also tested in practical applications. Companies will furthermore benefit from developing in-house training offers. The Austrian start-up mytalents.ai, for instance, offers AI workshops adapted to the tools used in a specific company. Martin Giesswein advises decision-makers to take the bull by its horns: “Discussing and playing around with AI with colleagues from various departments under the guidance of a trainer is an excellent way to familiarize oneself with this topic on a professional level.”

2. Two in One: Ethical AI Use and Compliance with the AI Act

This is a chance to tick two things off the list at once: ensuring ethical AI use while at the same time abiding by legal stipulations. “My recommendation is for managers to keep compliance with the AI Act in mind already now by drafting documentation for the classification that will become necessary in the future, whenever they plan and implement new systems,” Giesswein shares. In this context, it’s particularly important to...

• accumulate internal know-how and also rely on outside expertise.
• understand the AI Act’s risk framework (categorization into different risk levels).
• use the EU’s AI Pact, through which companies can voluntarily commit to implementing respective measures already now.
• keep up with the latest research on this topic.
• use standards and certifications for (new) IT systems with the aim of achieving compliance with the AI Act and ensuring ethical documentation, such as:
  • ISO/IEC/IEEE 7000: addresses ethical concerns during system design.
  • IEEE 7001 Transparency of Autonomous Systems.
  • CertifAIEd certificate and mark.

3. Draft Your Own AI Guideline

Voluntarily implementing a company-specific AI guideline already now makes for transparency and trust among employees, customers, suppliers, and other stakeholders if it answers the following questions:

• What is our stance on AI use?
• How do we use and monitor AI? What are our goals – downsizing the workforce through optimizing and automating processes or freeing up our employees to intensify human interaction with customers by letting AI handle routine tasks?
• How do we safeguard data protection and fairness when using AI?

“Only a handful of organizations already have an AI guideline, such as the Austria Press Agency or the City of Vienna,” Martin Giesswein shares about the situation in Austria. The Austria Press Agency (APA) looked into AI very early and introduced in-house guidelines for its use even before the launch of ChatGPT. Today, there is even an AI taskforce, and APA believes it is developing from a “trusted source of content” to a “trusted source of AI.” It also collaborates with other European media and technology enterprises, and in so doing acts as a potential role model for companies in other sectors. The City of Vienna has taken a different approach: a concise guideline of only two pages makes it transparent for both employees and citizens how AI is dealt with in the city government.

4. Estimate AI Use Impacts and Strategically Look Ahead

“Regardless of your plans for your company in the medium run, meaning in about three to five years, it’s very likely that AI will come to have a bearing. So make sure to account for it in your planning,” Stöttinger says. The following questions could be useful in this regard:

• Have new competitors or contestants from other industries entered your field of business as a result of AI use?
• Are customers’ demand patterns likely to change?
• Will the traditional ways of closing business deals be replaced by differed methods due to interactions with AI?
• Does the social change fueled by AI create new needs your company could cater to?

If CSR, ESG guidelines, and non-financial reporting are already assigned to certain departments in your company, staff should keep their eyes peeled for any relevant changes caused by AI.

5. Ensure Productivity Increases Across Your Organization

Proponents of using AI in daily business usually point to the potential of massive productivity increases. But technology by itself has rarely ever saved an organization a buck or two, much less created additional revenue. It’s only when technology is used in the right way that it can optimize productivity. The following steps have proved helpful in the past:

• Train managers and employees on how to use AI.
• Identify fields in which AI use makes sense: this could be optimizing logistics, forecasting, budget simulations, automating manual routine tasks in the office, marketing, or controlling interactions with agencies.
• Provide a company-specific AI environment adapted to your needs. Companies such as 506.ai based in Linz (Austria), Aleph-Alpha in Germany, and of course also Microsoft offer such solutions marketed as “CompanyGPT.”
• Make a long-term plan for increasing AI productivity. Also when it comes to AI, quick wins rarely last for long.

6. Stay in the Know with Regard to Your Country’s AI Act Implementation

“If the EU is to stay competitive, the countries’ respective AI offices must be closely aligned and ensure a uniform implementation of the AI Act across all of Europe,” Stöttinger says.

The AI Act and its strict rules are making the EU a pioneer, which European companies should use to their advantage. For companies to be able to make the most of this new regulatory reality, however, local authorities must act in unison and without excessive red tape. “The AI Act could create great opportunities for companies – or do the exact opposite. Even more reason to start getting ready: because there is no doubt that the AI Act will be passed,” Stöttinger concludes.

Update for Leaders

Join 15,000 + professionals and get regular updates on leadership and management topics. Learn something new every time.