Why AI is stagnating despite leaders’ optimism

As artificial intelligence (AI) increasingly establishes itself as an essential technology in the global economic landscape, a paradox persists in 2026: despite a clear enthusiasm from business leaders, its actual adoption remains far below expectations. Boards of directors, executive committees, and strategic forums constantly proclaim the crucial importance of AI for innovation and competitiveness. Yet, on the ground, the concrete implementation and operational deployment of AI projects struggle to progress. The gap between rhetoric and action raises deep questions about the hidden obstacles that slow down the digital transformation of organizations.

This paradoxical situation is due to several combined factors: insufficient project preparation, a glaring lack of methodology to measure impact, technical integration difficulties, as well as significant cultural resistance. Artificial intelligence initiatives are often launched under market and competitive pressure rather than from a solid strategic perspective. Thus, the optimism expressed by leaders is not enough to mask the profound limits of digital and organizational maturity within companies. Understanding these blockages is essential to envisage a sustainable and effective evolution of AI in business in the coming years.

The deep reasons behind the stagnation of artificial intelligence in business

Despite an almost unanimous recognition of AI’s importance – nearly 77% of strategic management in France place it at the top of their priorities – true integration remains largely hypothetical. Why such a gap? The main identified barrier lies in insufficient project preparation. Two out of three companies do not conduct a clear profitability analysis before launching an AI initiative. Without this rigorous evaluation, decisions often rest on poorly founded assumptions, making reliable management and optimal resource allocation difficult.

This lack of precise financial indicators directly affects the credibility of these projects. Indeed, the ability to demonstrate operational efficiency or generated savings is essential to sustain investments. A greater mistrust then appears towards AI initiatives, as tangible results are awaited. This phenomenon creates a vicious circle where doubt hinders acceleration, while investments do not receive the expected support within internal administrations.

Moreover, 94% of companies encounter difficulties industrializing AI technologies on a large scale. There is a multiplication of experiments and proofs of concept (POCs), but their transition to operational production mode remains limited. In this context, real advances in business processes are slow, even almost nonexistent. AI adoption is often perceived as a reaction to external competitive pressure rather than the result of a thoughtful strategy focused on value creation.

Data dispersion, fragile technical architectures, as well as the lack of alignment between different stakeholders (general management, business units, IT services) weigh heavily on the harmonious deployment of solutions. As Pete McEvoy, global head of AdvisoryX, points out, the absence of solid foundations compromises every attempt at industrialization, resulting in fragmented and unproductive use.

The critical role of governance and corporate culture in the face of leadership optimism

Beyond technological challenges, governance and organizational culture appear as determining factors in the dynamics of artificial intelligence. A recent study reports that 86% of companies face major cultural obstacles in AI adoption. Resistance to change, often underestimated, constitutes a real barrier to integrating new technologies.

Teams, confronted with the novelty of tools and sometimes fearing the transformation of their own roles, display a certain conservatism. This apprehension fuels a form of mistrust, heightened by the impression that AI could dehumanize processes or replace certain talents. The absence of appropriate training and the shortage of specialized skills further worsen this situation. Indeed, 23% of companies lament a lack of qualified talent to lead AI projects.

This cultural dynamic highlights the need for clear leadership, capable of uniting around a shared vision. It is also necessary to promote adapted governance, integrating risk management, transparency on data usage, and an ethical approach. Without a rigorous framework, the risk of derailments or failures increases, feeding decision-makers’ caution or skepticism.

Inclusive governance fostering collaboration among business units, management, and IT experts is the key to go beyond mere enthusiasm and initiate a deep and lasting transformation. Companies investing in this regard often observe better engagement from teams and a smoother adoption of artificial intelligence tools.

A list of the main cultural barriers to AI adoption

• Resistance to change and fear of disruption to jobs
• Lack of training adapted to new AI tools
• Shortage of specialized profiles and technical skills
• Absence of a shared strategic vision and clear governance
• Fear of social impacts including job transformations
• Lack of transparent communication on challenges and benefits

The challenge therefore consists in supporting skill development and involving employees in a participatory approach centered on the added value of AI in their daily tasks, so that this technological revolution is no longer seen as a threat but as an opportunity.

The deficit in financial and operational impact evaluation: a major limit for AI investments

One of the major issues contributing to the apparent stagnation of artificial intelligence lies in the weakness of measurement indicators. Very few companies succeed in accurately measuring the real impact of their AI projects. Financial returns are poorly defined, loosely quantified, or inadequately communicated, creating a form of opacity around the investments made.

This lack of rigorous evaluation affects overall strategy, due to a lack of reliable information on profitability. Companies thus miss the opportunity to optimize their project portfolio, prioritize those with high value-creation potential, and adapt their budgets with discernment. This phenomenon fuels the perception of a costly investment with uncertain benefits.

Setting up a clear set of indicators – financial, operational, and linked to adoption – appears as a sine qua non condition to revitalize the use of AI. Without these benchmarks, governance struggles to convince shareholders and stakeholders to continue down this path.

Type of indicators	Main objective	Concrete example
Financial indicators	Measure return on investment (ROI)	Reduction of operational costs through automation
Usage indicators	Evaluate the proportion of active users	Adoption rate of AI tools within teams
Business performance indicators	Analyze impact on productivity	Improvement in processing time of customer requests

Companies that manage to establish these measures are better able to align their actions with market expectations and work with a clearer development plan. This methodology helps strengthen confidence in the technology and encourages longer-term investment.

Competitive pressure: an insufficient driver for sustainable AI development

While competitive pressure remains a powerful driver for introducing artificial intelligence in companies, it is not enough to ensure lasting adoption. Many initiatives arise as strategic responses to external threats, customer demands, or progress made by direct competitors. However, this reactive dynamic may lack vision.

Companies engaged in AI under this impetus often lack deep reflection on the value truly generated. They multiply proofs of concept without guaranteeing that these experiments lead to lasting adaptation of business processes. This fragmented approach benefits neither growth nor long-term innovation.

The key lies in a reorientation toward thoughtful use, centered on transforming the company and creating a competitive advantage based on added value. Fidji Simo of OpenAI illustrates this point with a clear vision: it is a company that doubles its capabilities thanks to AI that will progress faster than one that simply tries to cut costs.

This philosophy invites considering AI not only as a performance lever but also as a catalyst for organizational innovation. To this end, it is imperative that external pressure is combined with a solid internal strategy.

Technical development and limits related to infrastructure and data

Technological foundations represent another major determinant in the stagnation of artificial intelligence. The massive deployment of AI solutions requires a solid architecture, capable of managing the complexity of data and algorithms. However, in many companies, existing systems are scattered, insufficiently integrated, and struggle to meet growing needs.

Data quality and governance constitute a fundamental challenge. Data often remains dispersed among different business silos, complicating its exploitation and harmonization. Without a unified data base, the relevance of AI models diminishes considerably, directly affecting result reliability.

IT teams must also manage infrastructure complexity, ensure cybersecurity, and guarantee real-time availability. These technical challenges, combined with a chronic shortage of specialized skills, slow projects down and generate additional costs that are often poorly anticipated.

It is therefore essential to deeply rethink digital infrastructures to remove these barriers. Simplifying tools and adopting modular and scalable platforms make a smoother and more stable AI adoption possible.

Example of a successful infrastructure reorganization

A large industrial company recently carried out a complete overhaul of its systems by implementing a centralized data platform coupled with embedded AI solutions. This transformation improved responsiveness in predictive maintenance, thereby saving several million euros annually and drastically reducing downtime.

Financial investment: a mixed reality despite leadership optimism

According to recent observations, nearly 68% of CEOs plan to increase their investments in artificial intelligence by 2027. This financial commitment sends a strong signal of confidence, but it is important to analyze the nature and effectiveness of these expenditures.

Allocated budgets do not always contribute to an overall maturity rise. A significant portion is consumed by repeated experiments which, without methodological support, struggle to generate concrete results. This dispersion reduces real market impact and gives the impression of a poorly structured investment.

The diversification of AI project types also raises the question of their coherence within the strategic roadmap. Leaders are advised to focus efforts on high-value-added solutions and avoid the temptation of “trendy” or mere technological gadgets.

In this context, better coordination between financial, technical, and human objectives is necessary. Investments must support a complete transformation integrating skills, governance, and technology.

The challenges of rebalancing skills in the AI-related labor market

Artificial intelligence profoundly changes skill requirements. In 2026, 81% of leaders anticipate a reshaping of the profiles needed to support these transformations. Traditional jobs evolve, new roles emerge, notably in data processing, cybersecurity, and specialized software development.

This evolution puts pressure on human resources. Recruiting specific talents becomes a major challenge while the market faces a significant shortage in these sectors. Concurrently, ongoing training of existing staff becomes a necessity to ensure harmonious integration of AI technologies.

Rebalancing must also consider human and social aspects linked to the massive adoption of artificial intelligence. Companies must anticipate changes, support transitions, and ensure the preservation of a climate of trust and sustainable engagement.

• Increased demand for expert data scientists and AI engineers
• Strengthening cybersecurity skills to secure AI systems
• Development of capabilities in agile project management and digital transformation
• Implementation of cross-functional training to facilitate understanding of AI challenges
• Promotion of new collaborative practices between business and IT

Future perspectives: towards a more mature integration of artificial intelligence

Despite current difficulties, the medium-term trajectory of artificial intelligence remains positive. The study shows that 36% of leaders plan to adopt more advanced and autonomous forms of AI, called “agentic,” within the next two years. This projection marks a clear willingness to expand use beyond simple automation tools.

To succeed in this stage, it will be necessary to surpass the chronic experimentation phase and embed AI at the heart of business processes in a sustainable and strategic manner. This involves revising business models and improving coordination among the company’s different spheres.

This evolution should also be accompanied by significant work on transparency, ethics, and data governance. Only responsible practices will ensure balanced development, beneficial both for companies, employees, and society as a whole.