- Scriptorium Initiative | Page Partner
Link Sharing: When AI Agents Become Your Communications Team | Scriptorium Initiative
April 22, 2026
Teams of autonomous AI agents may soon be able to handle complex communications challenges on their own. An experiment by the Scriptorium Initiative, a think tank exploring the future of communications in the AI era, suggests the transformation of the profession could be far more profound than most leaders expect.
The communications executives who participated in the experiment were stunned.
The question, says Nanne Bos, Chief Brand and Communications Officer at Aegon Group, is no longer whether AI will change the communications profession. The real question is how quickly and how fundamentally.
“I’ve always been fascinated by digital communication. But at a certain moment I realized we are approaching a threshold. A point where humans and technology come so close together that real interpersonal communication through systems becomes possible.”
What intrigues him most is the convergence of emotion and rationality.
“The idea that emotion and reason could merge in a conversation with a system fascinates me enormously.”
That curiosity led to a practical experiment inside Aegon.
The result was Luigi, an AI assistant trained on the entire historical output of the company’s corporate communications function.
“Luigi now helps write almost everything we produce. Annual reports, press releases, speeches. It works so well that Luigi already has an English and an American cousin. Soon we’ll add an asset-management variant trained on the communications of our investment business. Luigi keeps evolving.”
Bos quickly realized the implications extended far beyond one organization.
“AI is changing how we work, how we communicate, how we build and protect reputations. This topic is too disruptive to keep inside the company. People first need to understand what AI is, what it can do, and what it cannot do before they are willing to work with it.”
To explore these questions more broadly, Bos founded the Scriptorium Initiative together with brand consultant Marc Cloosterman and entrepreneur and copywriter Eric van Hall.
The initiative functions as a think tank where communications leaders examine how AI will reshape their profession.
“We created a space for reflection and experimentation. Our first step was a white paper outlining five predictions about the secondary effects of AI on communication. It sparked a lot of discussion.”
The initiative later hosted a conference featuring speakers from the University of Oxford and several future-oriented media research groups. But reflection alone was not enough. The founders wanted to experiment.
Bos became particularly interested in agentic AI. Systems capable of autonomous reasoning, planning and decision-making.
“An agentic system doesn’t just execute tasks. It can think along, advise, and even challenge your assumptions.”
The idea for the next experiment emerged during a conversation with Erik van der Loo, professor at INSEAD and expert on boardroom dynamics.
“He pointed out the growing role AI agents may play in decision-making. That triggered a thought. What if we built an AI board to advise the Scriptorium Initiative?”
The team first designed the structure of an ideal board. Number of seats, roles, diversity of perspectives. Then they mapped those roles to historical personalities.
Leonardo da Vinci and Steve Jobs represented innovation. Oprah Winfrey represented culture and empathy.
“These figures became the inspiration for AI personalities. Together they form an agentic board. They debate, disagree, complement each other, each bringing different perspectives.”
Once the board worked, Bos pushed the idea further.
“What if my entire communications team at Aegon consisted of AI agents?”
The team defined roles such as strategy, content creation, spokesperson, project management and ethics. Each role was linked to a distinct AI persona.
“That’s how we created a fully agentic communications team.”
For Bos, the exercise forced a fundamental question.
“What is the real purpose of a communications department?”
Not the tools. Not the tasks. The value it creates.
“I believe the essence of communication is building trust. Trust in an organization among all stakeholders.”
When trust becomes the guiding principle, an AI system makes very different decisions compared with a system optimized purely for production or persuasion.
The experiment became even more interesting when the agents were given visual identities.
“We literally gave them faces. That changed everything. As soon as an agent has a face, people interact with it differently. It starts to feel like a team.”
Voices were added later, allowing users to speak directly with the agents. Different AI models were connected to different roles through an automation platform.
“The social media agent runs on Grok and is trained on X data. That means it actually thinks in a social media context. Some agents can even access monitoring tools or social media accounts.”
To test the system, the researchers invited communications leaders to run a crisis simulation.
The scenario was based on a modern adaptation of the classic Tylenol case. In the exercise, Donald Trump attempts to warn about a chemical substance during a press conference but repeatedly mentions the brand name Tylenol dozens of times.
The result was surprising.
“Within an hour the AI team produced a complete crisis strategy. Immediate response, long-term messaging, influencer outreach, the full package.”
Experienced communications leaders were stunned by the speed and depth of the output. Bos believes many operational tasks in communications will soon shift to similar agentic AI teams.
“What remains are the truly strategic roles. People who can think conceptually, maintain perspective, and direct teams of AI agents.”
That shift creates a new challenge. Traditionally, communications professionals learn their craft through experience. Handling crises, making mistakes, learning from them.
“If AI takes over much of that operational work, how do people develop the experience required to guide AI systems effectively?”
In Bos’s model, a human always remains responsible for the final output.
“You cannot blame the system after the fact.”
Ethical oversight is also built into the workflow.
“We added an ‘ethical officer’ agent that reviews decisions. The process includes checkpoints and iterations. You start with multiple scenarios, refine them, combine them. Within thirty minutes you can produce a holding statement while still maintaining human oversight.”
One unexpected observation during the crisis simulation was psychological. A crisis normally feels intense. Physical pressure, adrenaline.
“With AI the process felt more distant. Some participants dismissed the results because it didn’t feel real.”
That, Bos argues, is a mistake.
“AI will also appear on the receiving side of communication. We will soon deal with synthetic stakeholders. AI agents representing individuals who filter and interpret information on their behalf.”
Another observation was that the AI team gradually became more corporate in tone.
“That is actually a risk. Communication systems must remain as human as possible.”
Sometimes the system should deliberately slow down before answering.
“A response that comes too quickly often feels unnatural.”
In the long term Bos expects fewer communications roles, but more fundamental ones.
“If communication is ultimately about trust, the nature of the work will change.”
Internal communication, for example, may eventually address not only employees but also AI systems within the organization. At the same time, human authenticity may become more valuable.
“In a world saturated with AI, people will increasingly crave imperfection and genuine emotion.”
Yet AI may become so convincing that distinguishing human from machine will no longer matter.
“What matters is whether the message resonates.”
AI can already replicate tone of voice with remarkable accuracy.
“At Aegon Luigi proves that. Tone-of-voice guidelines used to sit unused in documents. Now they can be embedded directly into the system.”
In the future, organizations may gain unprecedented control over communication.
These predictions originate from the Scriptorium Initiative white paper Communicating with Robots, Connecting to People.
Here's What's New and Why Your Input Counts Complete the 2026 Benchmark Survey Last year, senior com…
Page Society in collaboration with the Scriptorium Initiative The communications profession has reac…
In this member-submitted report, Aegon Chief Brand and Communications Officer Nanne Bos outlines a b…