What are the data and AI trends for 2025

What are the data and AI trends for 2025?

Let's take a look at the main trends and lessons learned: the rise of Generative AI, the simplification of deployment processes, and the challenges of AI adoption and acculturation.

Generative AI, the trend of the moment

Let's take a look at the main trends and lessons learned: the rise of Generative AI, the simplification of deployment processes, and the challenges of AI adoption and acculturation.

Generative AI continues to be at the centre of discussions, thanks in particular to its ability to create multimodal content and insights, and to transform the customer experience. This field, which has been the subject of numerous demonstrations at trade fairs, is attracting interest for its multiple applications and its potential impact in a variety of sectors.

Significant benefits for retail

With conversational agents and personalised recommendations, companies can offer an enhanced customer experience. Clarins, for example, has launched its 'Clara' chatbot, which has increased satisfaction by 40% while reducing returns by 30%.

The benefits of Generative AI for improving internal performance

From experimentation to industrialisation: the path to AI in business

Companies such as L'Oréal, PepsiCo, Fnac/Darty and Decathlon have shared their gradual transition from experimentation to the industrialisation of AI. These companies have made strategic choices, laid down clear rules, and launched their own AI Factory to structure the deployment of their AI initiatives. These key stages are evidence of a common path taken by major groups that have been able to adopt AI effectively.

This also involves simplifying the tools to achieve an effective ROI: AI solution publishers facilitate industrialisation by simplifying them so that they can be used by non-technical profiles. This makes ROI calculations more accurate and encourages large-scale adoption.

Multimodal AI and specific applications

With multimodal generative AI, companies can simultaneously exploit different types of data (text, image, audio) to enrich customer interactions and adapt their products and services.

The implementation of specialised agents for specific professions, or even generalist agents capable of assisting with a variety of tasks, reposition employees for higher added-value missions. The main challenges for companies remain optimising productivity and improving the employee experience.

Repositioning human resources

The deployment of these agents enables employees to concentrate on higher added-value activities, while repetitive or low added-value tasks are automated.

The fundamentals are still there, but with simplified implementation

Among the essential elements, data quality and governance remain major challenges for any artificial intelligence strategy. The rise of industrialisation processes, such as DataOps and AIOps practices, supports the deployment of AI and ensures smoother implementation, a key factor for companies seeking to accelerate their digital transformation.

In addition, AI-enhanced data platforms accelerate and automate the exploitation of strategic insights, while ensuring greater data reliability and quality at every stage of their value chain. This approach facilitates the activation of different business use cases and supports the transformation towards a truly data-driven enterprise.

Data governance, in particular, is essential for providing clear visibility of a company's data assets, their quality, uses and lifecycle. This information is essential for fuelling new uses of AI, identifying quality data, and ensuring compliance with regulations such as the RGPD.

We have also observed that the development of AI is progressing more rapidly in companies structured around data meshes. These companies are organised around product domains, which encourages business autonomy in the production of datasets, while maintaining a cross-functional technical, organisational and governance framework. This model facilitates the creation of data marketplaces, encouraging more open consumption of this data, as opposed to a 'product on demand' approach.

The challenges of acculturation and adoption of AI

A question of corporate maturity

The adoption of AI varies according to the maturity of each company, and the major groups recognise the importance of training their employees to support this transition. Danone, for example, has embarked on a massive training programme for its 90,000 employees, while Microsoft has launched an AI training platform, which aims to train 1 million French people in AI by 2027. These initiatives clearly illustrate the challenge of acculturation to AI.

Bring Your Own AI (BYOAI) and Shadow AI: with the emergence of the BYOAI concept, where employees integrate their own AI tools, businesses are faced with the risk of 'Shadow AI', which, like 'Shadow IT', is often beyond the control of the IT department. This phenomenon represents a challenge for governance and security.

A real HR challenge: how do you harmonise the work of experienced employees, who have been in post for 30 to 40 years, with that of the new generations who are integrating AI into all their processes? This difference in practices can cause difficulties in terms of collaboration and knowledge transfer. Companies urgently need to anticipate these HR challenges to ensure cohesion and productivity.

AI and regulations: towards legal adaptation by country

The legal dimension of AI remains a subject of debate. With the IA Act (European law on the use of AI), companies are trying to understand and anticipate the implications of these regulations. Each country is adapting its rules, which means compliance efforts for international groups and raises questions about how AI will be deployed in each region.

AI is rapidly being democratised, while revealing challenges in terms of governance, security and people. Companies are faced with decisive choices when it comes to structuring their adoption of AI and maximising their return on investment, while ensuring internal cohesion and complying with regulations. Over and above technological trends, the real challenge lies in the ability of organisations to support their employees in making optimal use of AI, to form a shared culture, and to prepare for the lasting transformations that this technological revolution will bring.

In this dynamic, the rise of agentic AI is opening up a new chapter: these autonomous agents, capable of performing complex tasks proactively and adaptively, are transforming the way we interact with digital systems. They no longer simply assist the user, but take the initiative, learn continuously and interact with other intelligences to optimise processes. This evolution is redefining collaboration between human and machine, while posing new challenges in terms of ethics, supervision and transparency of automated decisions.

Read more