Belgian Employee Harbors AI Suspicion, but Salesforce Counters with Trust and Training

Belgian Employee Harbors AI Suspicion, but Salesforce Counters with Trust and Training

Salesforce strongly believes in the AI-based transformation of businesses, and companies themselves seem at least very curious about what they can do with AI. Among employees in Belgium, there is still some fear, but according to Salesforce, this is not an unsolvable problem.

“AI Agents take care of repetitive tasks, but the future of work is very human. People and AI must work together.” This quote comes somewhat surprisingly from actor and Oscar winner Matthew McConaughey, who addresses the audience of Salesforce Innovation Day in Brussels via video message about the pleasures of AI. “AI, built with trust, elevates people.”

Suspicion and Potential

It’s not surprising that Salesforce has enlisted a familiar face to proclaim this message. Companies are on board with the story, but in the Belgian workplace, AI can still encounter necessary suspicion. Although Belgian employees struggle with questions about their own jobs and trust in AI, they also see the potential. This is evident from a survey conducted by iVox on behalf of Salesforce among 1,000 working Belgians.

A minority of 34 percent of respondents are clearly positive about the idea of AI in the workplace, and eight percent are enthusiastic. That’s not a lot, while Salesforce has fully embraced AI, and more specifically AI agents, since Dreamforce last year. The figure is in line with the results of a similar study a year ago.

AI in the Enterprise

The fact that seven out of ten Belgians are open to a digital colleague opens doors, which Salesforce is determinedly walking through. “Welcome to the agent-based enterprise,” says Lien Ceulemans, Country Director for Belgium and Luxembourg, enthusiastically on stage. She comes to tell a full hall in the Event Lounge near Brussels how Salesforce integrates AI in a reliable and practical way.

To this end, Salesforce recently announced several innovations that coexist under the name Agentforce 360. This platform offers the possibility to build and deploy agents at scale (and with the help of natural language). Via Data 360 – the successor to the Data Cloud – the agents are connected with data. Finally, the message is to deploy the AI colleagues for internal and external use.

Everything we release, we have already implemented ourselves in alpha and beta.

Gianni Cooreman, Senior Director Solution Engineering Salesforce

With Agentforce 360, Salesforce has a technologically conclusive story. “Moreover, we are customer zero,” says Gianni Cooreman, Senior Director Solution Engineering at the company. “Everything we release, we have already implemented ourselves in alpha and beta.” Cooreman and his colleagues also use AI agents themselves in their daily work for Salesforce, on the Salesforce platform.

Frightening

Just as important as the technical story, however, is the Workforce Innovation Playbook. The reluctance of Belgians towards AI largely comes from uncertainty about the impact of the technology on their job. 27 percent of respondents indicate that they find AI frightening.

“You have to take people’s concerns seriously,” Ceulemans says about this. “There are different ways to show that AI is something employees themselves can embrace. When they work with the technology themselves, they will see the added value in their own team.”

Lien Ceulemans, Country Leader Belgium & Luxembourg Salesforce

“That’s why it’s so important to show an overarching vision as a company,” Cooreman adds. “A clear strategy is essential. AI is not there to eliminate jobs. Make clear what the intention is, for which use cases AI is being deployed, and what the plans are with the productivity gains achieved.”

No Job Loss but Shift

The fear among Belgian employees is certainly there. 48 percent fear job loss, but according to Cooreman, this will not be the case. He is convinced of a net positive effect of AI on jobs. “Salesforce will have more employees at the end of this year than at the beginning,” he says, although he immediately adds that it is difficult to immediately estimate the real impact at the beginning of a new technology cycle.

Shifts are on the agenda. The Innovation Playbook describes important human pillars for successful integration of AI. “Reskilling is essential,” Cooreman knows. The job content is changing. At our company, everyone has been required to take the Trailhead course on AI. I myself am busy every day learning about the innovations.”

Ceulemans agrees. “Learning new skills is very important. We’ve always focused on that at Salesforce, even before the advent of generative AI.”

Learning new skills is very important. We’ve always focused on that at Salesforce.

Lien Ceulemans, Country Leader Belgium & Luxembourg Salesforce

Furthermore, companies need to think about how they can deploy the talent that becomes available in other parts of the organization. Internally, Salesforce uses the Career Connect tool for this, where employees are connected with new functions within the organization.

Useful, but also Trustworthy?

The surveyed Belgians already have a vision of how they want to use AI. 39 percent want to use AI agents as a knowledge source that can quickly provide answers. 37 percent see potential in an assistant that can take over routine tasks and administration. 33 percent believe in AI for creative work and 21 percent already see AI as a data analyst. According to Cooreman, this is a goal where there is a lot of potential.

Finally, the respondents point out a lack of trust. Almost half (49 percent) do not trust decisions made by AI agents. It’s notable that women are more suspicious of what AI says than men (43 percent vs. 56 percent). Especially younger employees indicate that they trust AI more.

This finding can be linked back to the core of Salesforce’s AI offering for the workplace. Trust has been central to this since its launch. “Yet training is also crucial,” Cooreman adds. “Not all AI functionality is suitable for all purposes. It’s important to know where AI can help and in what way.”

Strict Policy Rules

In the context of Agentforce, Salesforce provides trust in various ways. The Atlas Reasoning Engine, which interprets questions, plays a major role, as does the linking of the right data and retrieval augmented generation. The hybrid reasoning engine then adds a piece of deterministic rules (if X, then Y) to the reasoning capabilities of AI agents.

Governance is also important. For example, Salesforce enables policy management for unstructured data. On stage, we see a demo where policy rules for a chatbot are added in natural language. This way, the AI agent shares practical information from a large PDF with a human user, but the bot doesn’t reveal contact details of employees that are also in the PDF.

The Salesforce team in Belgium also strongly believes in the role of AI. The fact that employees are somewhat more reluctant illustrates the importance of adequate change management in every AI implementation. Many people are fearful or suspicious, and projects will only succeed if these employees also get on board with the story.

Measurable Added Value

From the business side, there is a lot of interest. The keynote during Innovation Day is certainly packed. Salesforce itself tries to set an example. “Salesforce’s customer support agent has already had 1.5 million conversations with customers,” McConaughey says in the video.

This yields results. Maarten Laukens, Deputy Country Leader: “With this, we have saved a hundred million dollars on an annual basis. Agents in sales have supported sixty million dollars in sales in the pipeline, and our conversion has improved by fifteen percent thanks to extra personalized communication. 68,000 employees work with AI agents in Slack.”

AI in a business context thus delivers measurable results. Guidance, training, and change management should get even reluctant Belgians on board, and then little stands in the way of an agent-based transformation of the enterprise.

“Every company in Belgium today knows that AI is a revolution,” Ceulemans concludes. “We’re past the experimental phase, due to the relatively low entry cost and the simplicity of the product. Organizations have agents in production, also with their customers. It works.”