Skip to content
Skjld Labs

How to manage the risk of unauthorised AI use in marketing

Marketing departments see a sharp rise in efficiency when employees start using artificial intelligence, but this development often creates an invisible security risk. When tools are adopted outside the IT department's control, sensitive customer data and trade secrets can end up on third-party platforms. For Norwegian small and medium-sized enterprises, the solution is about striking a balance between innovation and control.

3 min read Laget med AI-hjelp
Two people working on documents at a long wooden table in a bright room

The challenge of 'Shadow AI' in the marketing team

As tools like ChatGPT and Claude become universally available, we are seeing a growing trend of what is called 'Shadow AI'. This means that employees in marketing departments adopt AI tools to solve tasks without the organisation having approved or even seen the tools. The intention is often good. People want to make their workflow more efficient, but the consequences can be hard to keep track of for the business as a whole.

A review from singulr.ai describes the risk of employees processing sensitive customer data in these tools without authorisation. This can happen when customer lists are uploaded for analysis, or when personally identifiable information (PII) is used to draft text suggestions. The problem is not necessarily the technology itself, but the lack of oversight over where the data flows and how it is handled by third-party providers. When employees use free services without corporate agreements, you lose control over whether submitted data is used to train future models or whether it is stored in breach of internal security requirements.

When automation creates invisible security gaps

The risk is not limited to manual uploads in a browser. Modern marketing increasingly relies on automated workflows via platforms such as Zapier or n8n. These tools can connect different services and move data automatically between systems to save time.

If such a workflow is set up without central control, sensitive financial data or strategic plans can be moved between different AI services and marketing tools without those responsible for security knowing about it. This creates a gap between the established IT guidelines and the actual practice taking place in the marketing department. Without insight into which tools are actually in use, it becomes impossible to enforce data protection and compliance with privacy regulations such as the GDPR. Automation means that leaks can happen at a pace and volume that is hard to detect manually, which increases the vulnerability of the business.

Controlled acceleration rather than prohibition

For leaders in small and medium-sized enterprises, the temptation can be strong to impose strict bans to secure the data. Experience shows, however, that such bans often lead employees to simply move their usage to even more hidden channels, which further reduces the organisation's overall oversight. Instead, you should consider an approach based on what is described as 'controlled acceleration'.

This means implementing governance frameworks that give the organisation real-time visibility. Rather than stopping the use of AI, you should make it possible for employees to use the best tools within safe boundaries. This requires moving from static policy documents to active security measures that can catch risk before the data leaves the organisation's control. By creating a culture of safe use, you can reap the benefits of the technology without sacrificing control over critical assets.

Practical steps for Norwegian SME leaders

To build a secure and effective AI strategy, Norwegian businesses should focus on several core areas. First, you need to get a real overview of which AI services are actually in use in the organisation. This includes both the common chat interfaces and the underlying integrations in marketing tools. Without an overview of the actual tools, security work remains reactive.

Second, it is important to understand how the various providers handle data. It is critical to note that whether an AI provider stores or uses submitted data to train its own models often depends on the account type you have and the settings you have chosen. Businesses should therefore prioritise professional solutions with clear data processing agreements and use enterprise versions of the tools, where the privacy settings are stricter than in free services.

Third, you must integrate security directly into the automated processes. By setting up barriers that prevent the transfer of sensitive data in workflows, you can maintain high efficiency without compromising privacy. This means looking at the entire chain of data flow, from the CRM system through the automation tool to the AI model. By combining technical control with clear guidelines and continuous monitoring, you can ensure that the technology becomes a resource that supports growth, rather than an unpredictable risk to the company's reputation and data.

From article to practice

Want to put this to work?

Local and Norwegian AI. By default, the solutions we build run on our own servers in Sandefjord. If your data is sensitive, we set them up on site at your premises.

Read about Local and Norwegian AI Book a free call
Safe and local AI

Read on.

Newsletter

AI in practice, one email a month.

What we've tested, one tip you can use, and what's coming next.

You'll get the Skjld Labs newsletter once a month, and can unsubscribe at any time. Privacy

First step

Thirty minutes. Free.

A no-obligation conversation about what you are facing. We will say so if AI is not the right answer. We are happy to meet you at Langes gate 8 in Sandefjord, or on video.