\n
Artificial intelligence (AI) is no longer a marginal technology. By 2025, it will be transforming businesses, professions and organizations. Yet adoption remains heterogeneous: some companies are barely experimenting, while others are building AI-centric organizations. This dossier summarizes the state of play and best practices for adopting AI responsibly, based on data and studies published in 2025. <\/p>\n
Canada (Quebec).<\/strong> A study by the Institut de la statistique du Qu\u00e9bec summarizes the use of AI in companies in 2025. In the second quarter of 2025, 12.7% of Quebec companies were using AI applications for production<\/strong>, a proportion close to that of Ontario (13.3%). The increase remains modest compared to 2024. Sector analysis shows that finance, information, culture and professional services have adoption rates between 36.9% and 55.0%<\/strong>. The least advanced sectors are agriculture (0.3%), accommodation and catering (1.9%) and construction (2.3%). Size also influences adoption: large companies (\u2265100 employees) have a rate of 26.1% versus 12.2% for very small companies<\/strong>. <\/p>\n<\/li>\n France (retailers and SMEs).<\/strong> The guide from the Direction G\u00e9n\u00e9rale des Entreprises (DGE) states that AI enables retailers to improve product recommendation, deploy chatbots, forecast demand, generate marketing content and detect fraud. Use cases presented include customer support, content creation, recommendation personalization and inventory optimization. A white paper published in September 2025 points out that 32% of French SMEs are using AI, and 58% of managers consider its adoption a survival issue<\/strong>. These figures show that small businesses are catching up fast. <\/p>\n<\/li>\n Small businesses and international SMEs.<\/strong> According to an article dedicated to SMEs, 75% of small businesses are already using AI solutions<\/strong>, and over 90% recognize the time and cost savings. However, 90% of small businesses say they have encountered obstacles<\/strong> when integrating AI. Among the difficulties, 47% mention not knowing how to use the tools, 31% cite incompatibility with existing systems and 26% mention security and confidentiality issues. <\/p>\n<\/li>\n World (Global Surveys).<\/strong> McKinsey’s 2025 report shows that 88% of organizations are using AI in at least one function<\/strong>, up from 78% a year earlier, but almost two-thirds have not yet begun to deploy AI enterprise-wide<\/strong>. Interest in AI agents is high: 62% of respondents say they are experimenting with agents, and 23% are deploying them on a large scale. Despite this adoption, only 39% of respondents are seeing an impact on the bottom line<\/strong>, reflecting a gap between experimentation and value creation. <\/p>\n<\/li>\n<\/ul>\n Canada (KPMG).<\/strong> A survey conducted in August 2025 shows that 51% of adults use generative AI at work<\/strong>, with the majority (79%) observing productivity gains. More than half of users claim to save between one and five hours a week. However, 83% want or need training to use AI more effectively<\/strong>, and less than half (48%) feel their organization provides adequate support. Although 93% of executives say they are adopting generative AI, only 2% are seeing a return on investment<\/strong>. <\/p>\n<\/li>\n Marketing and communication.<\/strong> A survey of marketing decision-makers conducted by Semji reveals that 83% of companies have already integrated generative AI into their marketing strategy<\/strong>. However, 66% are still in the testing phase, and only 17% have deployed it on a large scale<\/strong>. Just 13% have not integrated AI, and less than 3% have no plans to do so. The main driver for use is productivity: 75% of professionals cite productivity gains, 64% automated content generation and 39% analysis time savings. <\/p>\n<\/li>\n Rapid change in adoption.<\/strong> A roadmap published in March 2025 finds that 31% of companies are using generative AI at the start of 2025, compared with just 12% at the end of 2023<\/strong>. The authors stress the need to define a clear vision and align AI with corporate strategy. <\/p>\n<\/li>\n<\/ul>\n The main AI applications adopted by companies cover several areas:<\/p>\n Text analysis and natural language processing.<\/strong> The Quebec survey shows that text analysis is the most common application: 56.7% of companies use it. Natural language processing (28.9%) and marketing automation (22%) follow. Other applications include data analysis (19.0%), conversational agents (17.9%) and content recommendation systems. <\/p>\n<\/li>\n Recommendations, chatbots and inventory management.<\/strong> The DGE guide lists concrete cases for retailers: personalized recommendation systems, chatbots available 24 hours a day, demand forecasting and inventory management, automated generation of marketing content and fraud detection. These solutions improve customer experience and decision-making. <\/p>\n<\/li>\n Productivity and automation.<\/strong> In marketing, productivity gains (75%) and automated content generation (64%) are the main drivers. AI is also used for personalization (30%), idea generation (28%) and SEO optimization (14%). <\/p>\n<\/li>\n<\/ul>\n The 2025 studies describe several technological and societal trends:<\/p>\n Mass adoption of generative AI and integration into business processes.<\/strong> According to Keyrus, generative AI will become commonplace and will be integrated into business processes, accompanied by multi-agent applications.<\/p>\n<\/li>\n Governance and ethics.<\/strong> Organizations will need to develop responsible AI, strengthen data literacy and integrate sustainability into data management. Trust and transparency are becoming essential criteria for AI acceptance. <\/p>\n<\/li>\n Infrastructure and technologies.<\/strong> Digital twins, AI for smart cities and the extension of cloud platforms are among the main technological levers.<\/p>\n<\/li>\n Sector-specific applications.<\/strong> AI is expanding into healthcare, education, cybersecurity, agriculture and industry. Opportunities include data monetization, AI for sustainability and advances in explainable AI. <\/p>\n<\/li>\n The evolution of AI agents and multimodality.<\/strong> A strategic guide describes the emergence of autonomous AI agents capable of planning, reasoning and collaborating with other agents. Multimodal systems that can process text, images, audio and video simultaneously should become widespread. <\/p>\n<\/li>\n Hyper-personalization and conversation.<\/strong> AI enables advanced personalization of offers according to the customer’s mood, environment or values, and conversational interfaces become a preferred channel. Sustainability and ethics are at the heart of innovations to reduce the environmental impact of models and guarantee algorithmic justice. <\/p>\n<\/li>\n<\/ul>\n Despite the enthusiasm for AI, many obstacles remain:<\/p>\n Lack of governance and framework.<\/strong> In marketing, 60% of companies have no formal governance framework for AI. The report highlights the difficulty of measuring ROI (46%), the complexity of integration (37%) and the lack of AI culture (34%). <\/p>\n<\/li>\n Skills shortages and training.<\/strong> The majority of Canadian workers want or need to learn about generative AI. Companies are struggling to offer adequate training programs. The SME article points out that 47% of small businesses don’t know how to use AI tools, and 31% encounter incompatibility with their systems. <\/p>\n<\/li>\n Security, privacy and bias.<\/strong> Legal and ethical concerns are central: 26% of SMEs cite data security and confidentiality as obstacles. Examples of bias and hallucination remind us of the need to control models to avoid discrimination or misinformation. <\/p>\n<\/li>\n Infrastructure and data.<\/strong> SMEs often lack the infrastructure to collect and process data, and lack structured processes for data quality and management.<\/p>\n<\/li>\n Gap between adoption and value.<\/strong> Globally, most organizations are using AI, but have yet to see significant financial benefits. Only 2% of Canadian executives report a return on investment for generative AI. <\/p>\n<\/li>\n<\/ul>\n To maximize the benefits of AI and limit the risks, studies suggest several approaches.<\/p>\n Team commitment and user experience.<\/strong> Expert Iga\u00ebl Derrida points out that nine out of ten AI projects fail mainly because of indifferent teams<\/strong>. He recommends creating functional prototypes (MVPs) and developing “lovable” products (MLPs<\/strong> ) that generate buy-in through careful design and a convincing user experience. AI needs to be seen as a business project, not just an IT one. <\/p>\n<\/li>\n Culture and acculturation.<\/strong> A roadmap proposes raising awareness among managers and teams<\/strong>: understanding AI concepts, aligning AI with strategy and defining clear objectives. Immersive, gamified sessions help to overcome resistance and get managers involved. Internal champions or ambassadors play a key role in spreading the AI culture. <\/p>\n<\/li>\n<\/ul>\n Align AI with strategy.<\/strong> Experts stress the importance of not doing AI for AI’s sake<\/strong>, and of aligning projects with overall corporate objectives. Time-consuming or repetitive processes need to be identified and the company’s digital maturity assessed. <\/p>\n<\/li>\n Start with targeted, measurable cases.<\/strong> The DGE guide suggests simple use cases for retailers (customer service, content generation, recommendations, inventory management). For SMEs, it is advisable to evaluate potential use cases in terms of added value and available data. <\/p>\n<\/li>\n<\/ul>\n Governance framework.<\/strong> The absence of a framework is a major obstacle. Organizations need to define governance policies, oversight committees and adoption metrics. Welcome to the Jungle details a model based on an AI transformation team, a program of “AI champions”, a training plan, rigorous monitoring of adoption and the development of tailor-made solutions<\/strong>. <\/p>\n<\/li>\n Responsibility and compliance.<\/strong> The DGE guide reminds us of the importance of anticipating the technical, ethical and legal issues associated with AI. RGPD checklists and regular audits are advised to ensure project compliance and protect customer data. It’s also crucial to assess biases, check datasets and monitor for hallucinations. <\/p>\n<\/li>\n<\/ul>\n Ongoing training.<\/strong> Surveys show that 83% of employees want training<\/strong> to master AI tools. Organizations need to set up targeted training programs and upskilling paths. The white paper for SMEs recommends combining incentives (bonuses, badges) to value the commitment of ambassadors and encourage a positive dynamic. <\/p>\n<\/li>\n Access to tools and infrastructure.<\/strong> SMEs need to assess existing systems, data quality and compatibility with AI tools. Where infrastructure is lacking, they can resort to cloud solutions or partnerships. Success depends on structured data collection and the use of accessible tools such as chatbots and SaaS platforms. <\/p>\n<\/li>\n<\/ul>\n Success indicators.<\/strong> Welcome to the Jungle measures adoption via monthly usage (number of messages generated), weekly usage, number of agents created and user distribution<\/strong>. These indicators are used to track the progress of the transformation and identify obstacles. <\/p>\n<\/li>\n Return on investment.<\/strong> Organizations need to assess productivity gains, financial impact and user satisfaction. Only accurate measurements will enable the transition from experimentation to value creation. <\/p>\n<\/li>\n<\/ul>\n The year 2025 marks a turning point: AI is no longer a luxury but a strategic lever for businesses. Statistics show a rapid progression in adoption, with variations according to size, sector and country. Generative AI is making inroads into organizations, but the majority remain in the experimentation phase, and the search for value remains a challenge. The companies that will succeed are those that align AI with their strategy, engage their people, structure responsible governance and invest in training. Transformation through AI is as much cultural as technological; it requires a clear vision, concrete use cases and impeccable ethics. <\/p>\n","protected":false},"excerpt":{"rendered":" Artificial intelligence adoption: state of play and strategy (2025) Artificial intelligence (AI) is no longer a marginal technology. By 2025, it will be transforming businesses, professions and organizations. Yet adoption remains heterogeneous: some companies are barely experimenting, while others are building AI-centric organizations. This dossier summarizes the state of play and best practices for adopting […]<\/p>\n","protected":false},"author":5,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"categories":[78],"tags":[],"class_list":["post-4851","post","type-post","status-publish","format-standard","hentry","category-artificial-intelligence"],"acf":[],"yoast_head":"\nAdopting generative AI<\/h3>\n
\n
Applications and adoption drivers<\/h2>\n
\n
Emerging trends for 2025<\/h2>\n
\n
Challenges and obstacles<\/h2>\n
\n
Best practices and adoption strategy<\/h2>\n
1. Putting people first<\/h3>\n
\n
2. Define strategy and use cases<\/h3>\n
\n
3. Structuring governance and ethics<\/h3>\n
\n
4. Investing in training and skills<\/h3>\n
\n
5. Measure and adjust<\/h3>\n
\n
Conclusion<\/h2>\n