Every year, Arctus’ Intranet Observatory provides an overview of the state of internal digital technology in French organizations. The 2025 edition confirms that the intranet remains a pillar of the work environment, deemed “indispensable or necessary” by 94% of respondents. However, behind these unsurprising figures, there are signs of a slowdown: intranets are becoming less customized and increasingly reliant on the Microsoft 365 ecosystem, and there is a growing gap between available features and actual usage.

But what interests me most in this edition is the emphasis on generative AI, which is also reflected in the change of name of the report, now entitled “Augmented Intranet Observatory.” In a way, the intranet is becoming a prism through which we can observe how businesses are trying to integrate these new capabilities into their tools, practices, and cultures, and the trend seems to be less toward revolution and more toward transposing old ways of thinking into new technologies.

In short:

• The intranet remains central to organizations, but is becoming more standardized around Microsoft 365, with declining customization and a growing gap between available features and actual usage.
• It retains a predominantly “info-com” logic, designed as a top-down information medium rather than a true integrated workspace, with low adoption of business and engagement features.
• Intranets are often deployed without a clear adoption strategy or governance, which limits their effective use despite their rich features.
• Generative AI is generating interest, but its integration remains superficial, with no transformation of processes or real consideration of the organizational, ethical, or economic impacts.
• Unconnected populations (such as field workers) remain excluded from internal digitalization, revealing a lack of inclusivity and a shared vision of the expected value of these tools.

An omnipresent intranet, but stuck in an “info-com” model

The first lesson to be learned is structural: the intranet remains a medium rather than a tool. In 40% of cases, it is designed as a top-down information system, often grafted onto SharePoint (58%), sometimes reinforced by an overlay or a specialized intranet CMS.

Digital Workplace and Employee Experience Platforms (EXP) are gaining ground (+2 points), but remain in the minority. We are therefore still a long way from a truly integrated workspace that is tailored to business uses or service-oriented. The intranet often remains the business’s home screen, but rarely serves as an operational hub.

This discrepancy is reflected in usage: basic functions (news, directory, EDM) are available but rarely used. Business functions (onboarding, workflows, training) are present in less than half of organizations. And engagement functions (recognition, ambassadors, gamification) remain anecdotal.

This observation disappoints me, but it doesn’t surprise me. I said the other day that the EXP trend had faded as quickly as it had appeared (Employee Experience Platforms: big promise, little delivery?) and for the rest, I have always found businesses’ ambitions in terms of intranets to be more than modest in relation to their needs, not to mention the digital workplace.

When this topic was at the heart of my business and my concerns, I argued for an intranet as a place to work and a place to live, because the limitations of an intranet primarily focused on communication are quite obvious: it does not meet the needs of users who, let’s not forget, are there to work, and some managers will even go so far as to say that it distracts employees by trying at all costs to keep internal communication alive.

So no, I’m not surprised. Some time ago, I read an article by the Digital Workplace Group (Modern intranets, A practical guide) and thought to myself that nothing had changed since the 2010s and what I and others had been writing on the subject. But this is by no means a criticism of the DWG’s vision, which is excellent, but simply an observation that businesses have lagged behind and that virtually none of the guidelines that experts were pushing 15 years ago have been followed through, if they were even considered in the first place.

Intranets deployed without an adoption strategy

The problem is well known: the most feature-rich applications are those that require the most support, and are therefore the least used. The study clearly shows that it is not the tools that are lacking, but organizational intent.

There is a real undersizing of what is known as “usage support”, which in fact masks a deficiency in organizational design. Who is responsible for usage? What are the indicators of success? What is the governance between IT, internal communications, HR, and business lines? How can work be transformed to take advantage of technology? In the absence of answers, the intranet, however rich it may be from a functional point of view, is an empty shell in terms of usage.

Here we find a constant feature of internal digital projects: the illusion of deployment. The functionality is available, so the service is provided, but the challenge lies elsewhere: in workflows, routines, and daily trade-offs between emails, tasks, instant messages, business tools, and the intranet (Digital Infobesity: When Collaboration Tools Degrade Productivity, QWL and Amplify Mental Workload).

A few months ago, practitioners in the field told me that compared to “before,” they were negatively surprised by the fact that businesses were investing less and less in support. I see two reasons for this.

The first is not new and is the overestimation of employees’ digital skills in a work context. Just because everyone knows how to use Facebook does not mean they know how to apply these skills in a professional context.

The second is another factor mentioned in the report, namely Microsoft’s overwhelming dominance in the sector. The question of which technology to choose no longer arises, so we no longer have to deal with it and suffer the consequences of disappointing adoption.

I would also like to draw a parallel with another topic, employee experience, of which the intranet is arguably a component. I won’t dig out the numerous articles in which I explain that the employee experience is primarily found in operations, but I did recently come across an article that explained very well that most employee experience initiatives did not know how (or did not want) to measure their business impact (How I Stumped a Panel of EX Experts).

In fact, apart from the fact that the intranet is part of the employee experience (but we’ve seen what happened to employee experience platforms), both suffer from a similar problem. Both should focus primarily on how work is done, whereas the employee experience has been hijacked by QWL and the intranet by communication.

Generative AI: a promise to be fulfilled

Generative AI makes a notable appearance in the 2025 study. 10% of respondents already use it on their intranet, 17% are in the process of deploying it, and 23% are in the evaluation phase. That means more than 50% are engaged in some way.

The two dominant use cases are content generation (58%) and augmented search (RAG, also 58%). These are classic use cases, but they remain very focused on contributors (to produce more) and users (to find things faster).

The problem is that we are applying a very traditional logic to AI: producing content and facilitating access to information, but without thinking about reconfiguring workflows, responsibilities, or cultural impact. We are a long way from Moderna… (HR and IT merger: Moderna redesigns its organization for and with AI) AI is becoming an assistant to do what we already did faster, without questioning the meaning of what we do.

The other blind spot is governance: cybersecurity, ethical frameworks, prompt training, structured data management… These dimensions are still largely overlooked, even though they determine the credibility and sustainability of projects.

I am really curious to see how things will evolve in the future. I am not going to revisit the enormous potential of AI, which no one can dispute, but I think that its adoption in the workplace will be anything but simple.

Firstly, because AI in business is not the same as consumer AI and faces constraints that cannot be ignored (Why enterprise AI can’t keep up with consumer AI: beyond ChatGPT, a more complex reality).

Then there is the “culture clash” between a deterministic business and AI, which is not (Businesses are deterministic, generative AI is not, and that’s a real problem.).

There is also the question of ROI. What we are seeing today is that businesses are struggling to tame AI (AI in the digital workplace: a brilliant assistant, but an unreliable colleague) but, above all, that the costs are far too high to simply generate content. This explains, in my opinion, Microsoft’s difficulties in selling Copilot: while 60% of businesses have tested Copilot, only 16% have moved on to the deployment phase (How to get Microsoft 365 Copilot beyond the pilot stage).

An IT professional at a large French company recently told me the same thing: with more than 100,000 employees, they decided to limit themselves to 5,000 Copilot licenses due to the cost.

Finally, AI is limited by the same factors that have prevented the creation of ambitious digital workplaces, and there is a risk that the same causes will have the same effects (Digital workplace, AI, and interoperability: a problem that remains unresolved).

The big absentees: the disconnected, the professions, and value

“Deskless workers” represent a significant portion of the workforce, but remain the forgotten ones of internal digitalization (Digital Transformation : don’t forget blue collars). With difficult mobile access and a lack of digital support, this problem, which was already identified in 2023, seems to have been completely overlooked, and no progress has been made in this area.

The budget priority has been AI at all costs, even if it means leaving behind an entire population that seems to be considered “underemployed.”

This is a sign of an intranet that is still designed from headquarters for white-collar workers with a “digital office” logic that is not very inclusive.

But it is also a sign that the expected value is not always clearly defined. What are we trying to improve? Communication? Efficiency? The employee experience? Engagement? Inclusivity? Business management? All of the above?

Bottom line

The Arctus 2025 study has the merit of stating the obvious: the intranet has not disappeared, but it has not yet reinvented itself either. AI could be a catalyst for transformation, but it will not replace the need for a clear strategy, shared governance, and a real understanding of business practices.

The intranet of tomorrow will not be based solely on technical overlays or conversational assistants, but will build on the ability to think of tools as work and action devices at the individual and collective level, adaptable, useful, and aligned with contemporary forms of work organization.

And it is perhaps this transformation that will lead us to an augmented intranet, which is still far from reality today.

Image credit: Image generated by artificial intelligence via ChatGPT (OpenAI)

Related articles:

Digital workplace, AI, and interoperability: a problem that remains unresolved Generative AI in the enterprise: a silo breaker or just another layer in an already complex IT landscape? Generative AI in the workplace: revolution or illusion? AI in the digital workplace: a brilliant assistant, but an unreliable colleague Employee Experience Platforms: big promise, little delivery? Digital Workplace and Intranet : the 2013 trends (part #2)