AI in communications – a personal interim status

Remaining touched

Between hype, high investment and disillusionment, one thing is clear: Artificial intelligence can efficiently support research, analysis, and idea generation. Quality, credibility, and transparency, however, are not optional values. Communication still requires thinking.

AI fever between hype and disillusionment

‘AI fever’ is the Swiss financial sector’s word of the year. Fever is an apt metaphor for the current experimentation, discovery, and implementation of systems based on machine intelligence. In fact, we are still very much in the early stages of this revolutionary new technology and in the midst of the hype phase – with all the uncertainties that this entails, for instance, in terms of security.

Gartner already sees the development of artificial intelligence for 2025 as being in the trough of disillusionment,¹ while the agency for global trend analysis states for 2026 that digitalization is no longer an option but a strategic necessity.² In addition, the question about business value of investments in AI-driven systems is now clearly emerging.

Despite average global spending of $1.9 million on GenAI initiatives in 2024, less than 30 per cent of AI leaders say their CEOs are satisfied with the return on AI investments, according to Gartner. Another challenge is finding qualified specialists and imparting the necessary GenAI knowledge internally.¹

So it's time to take a personal interim stock of communication projects and text creation with regard to usefulness, transparency, and data protection.

Pragmatic and responsible

First of all, learning, experiencing, and evaluating will continue to accompany us in the future and remain part of our regular tasks. Healthy curiosity is as much a part of our professional understanding as openness and willingness to acquire new knowledge, question established practices and adapt processes so that they are consistent with our values and professional ethics.

As a general rule, not everything new is automatically better, and not everything that has proven itself in the past is necessarily outdated. It is different. And with AI in particular, some things are changing, and in some industries and areas, the changes can be significant. This makes it all the more important to keep a cool head and act responsibly.

In the day-to-day running of my communications agency, as I experience it as a self-employed entrepreneur, the use of AI tools depends on the job at hand. Conducting interviews for magazine articles, for example, still takes place through direct interaction and requires sensitive conversation skills. Media relations are also based on human contact and mutual dialogue. Similarly, project management requires a continuous exchange of information that provides an overview in addition to brief digitally exchanged status updates. This is the only way to ensure that all aspects are taken into account and that there is consensus on the steps and responsibilities that will follow.

Preparation remains crucial. The path to this can involve research via the internet and AI platforms. I generate inspiration for possible questions using AI tools such as Neuroflash or ChatGPT. I then test the supplemented and adapted questions to get an impression of whether they sufficiently cover the readers' perspective. This initial feedback loop helps me to critically reflect on the questions once again and check their relevance and comprehensibility. However, the editorial work based on the interviews conducted remains entirely in my hands. The same applies to writing press releases or organizing media events.

Writing remains intellectual work

Writing means condensing content. It involves classifying, analysing, boiling down and distilling the essence. Writing remains intellectual work. It also means taking into account the reader's level of knowledge. A good text remains committed to quality.

The quality standards demanded of the media can be applied to institutional communication, especially since credibility is a key factor in a company's reputation. According to the second study on media quality conducted by fög, Research Centre for the Public Sphere and Society, the majority of the public surveyed considers it sensible to use AI primarily for translations, detecting false information on social media, evaluating data or for research purposes, but not for creating journalistic content.³ In its October 2025 guidelines, the German Public Relations Council (DRPR) states that PR and communications professionals are always required to label AI-generated content when it is created and published without being checked.⁴ The principle of accuracy and truthfulness also applies without restriction to work with AI tools.

AI in everyday communication – proven guidelines

The following guidelines for daily use can be derived from all of this. They are based on the four phases of the Com-Circle concept cycle (conception – creation – implementation – evaluation):

• Briefing meeting: Direct communication with the client remains the fundamental element throughout the entire communication process. Questions can be asked, and a shared understanding of the initial situation and objectives can be reviewed and agreed upon. This is most effectively and efficiently achieved through face-to-face communication.
  
• Research, analysis, brainstorming: Search engines such as the French Qwant, which promises comprehensive data protection, as well as AI tools such as Le Chat from Mistral AI, ChatGPT from OpenAI, Copilot from Microsoft or Gemini from Google provide support for initial rapid information processing. In-depth analysis and verification remain the responsibility of the agency and can be time-consuming, depending on the issue and area in question. In addition, the data protection guidelines agreed with the customer must be followed.
  
• Proofreading texts: Once the content has been developed, readability, style and dialogue group-oriented language can be checked using AI platforms, and variants can also be obtained for specific channels and media. These recommendations must then be evaluated by the agency. The agency remains responsible for the accuracy and quality of the content. Experience, editorial competence, awareness of tone, language style and literary rhetoric, as well as in-depth industry knowledge continue to characterize the expertise of copywriters.
  
• Evaluation and adaptation: Evaluation covers a broad field, ranging from media response analyses and social media sentiment evaluations to measurements of other communication channels and entire campaigns, both analogue and digital. Depending on their focus, providers have been working with AI-supported tools for some time now.

Food for thought on AI with a view to sustainability

What was that about the issue of resources? Shouldn't we as a society be able to consider development potential in its entirety, in both its positive and negative dimensions? How much energy is proportionate when a complex infrastructure such as ChatGPT is put into operation for a simple email text query?

What makes economic, ecological, social and ethical sense? This is a question that will hopefully be discussed in depth. And perhaps it is also a topic for the Fridays for Future movement!

References

¹  Gartner: ‘Key takeaways from the Hype Cycle for Artificial Intelligence’,The 2025 Hype Cycle for Artificial Intelligence Goes Beyond GenAI

² Gartner: ‘The 10 technology trends that will shape the next five years’,
Die wichtigsten Technologietrends 2026 | Gartner

³ Vogler D. et al, fög, Yearbook Quality of Media Study 2/2024, ‘Artificial intelligence in journalism: How acceptance is developing among the Swiss population’,
JB_2024_Studie_II_KI_im_Journalismus_final.pdf

⁴ German Public Relations Council, DRPR-Richtlinie zum Einsatz von KI in der PR - DRPR - Deutscher Rat für Public Relations

Further links

• USZ,  DSI (Digital Society Initiative), Director and Professor of German Linguistics, Noah Bubenhofer. ‘Why you shouldn't always leave writing to AI’, DSI Spotlights | Digital Society Initiative | UZH
  
• Centre for Digital Responsibility, Think Tank for Digital Ethics by Cornelia Diethelm, Centre for Digital Responsibility (Digitale Ethik)
  
• Finews ‘This is the financial word of the year’, Dies ist das Finanzwort des Jahres

Author

Liliane Elspass