Towards a responsible adoption of AI

The integration of artificial intelligence (AI) into our daily lives—whether in our private, professional, or educational lives—is experiencing an unprecedented pace of adoption. However, this growing ubiquity of AI is accompanied by a relational complexity marked by a profound ambivalence: between fascination with its benefits and concern about its risks and impacts.

Today, two-thirds of individuals say they intentionally use AI on a regular basis. This figure is indicative of the speed with which this technology has penetrated all levels of society. Yet, paradoxically, half of users admit to not having received any specific AI training, which raises essential questions in terms of digital literacy and individual responsibility in the use of these powerful tools.

The gap in adoption between emerging and advanced economies is particularly notable. While the former display high usage rates, often exceeding 80%, accompanied by an overall optimistic perception, advanced economies are more reserved in both their confidence and acceptance. This difference is partly explained by the relative benefits that users perceive in each socio-economic context: for emerging economies, AI represents a valuable opportunity to address certain structural deficits, particularly in terms of access to information, health, or education.

Despite this widespread adoption, trust remains fragile. More than half of individuals report having reservations about the security, privacy, and ethics of AI systems, even though they recognize their technical effectiveness and practical utility in everyday life. Thus, trust in AI remains deeply contextual and varies considerably depending on the intended application. For example, the use of AI in the medical sector generally elicits higher trust, due to the tangible and immediate benefits perceived by users.

This emotional ambivalence toward AI is particularly marked by a coexistence of optimism and concern. While excitement about the possibilities offered by AI is real, growing anxiety about its side effects, such as misinformation, the loss of authentic human interaction, and cybersecurity risks, is also palpable. The phenomenon of AI-generated disinformation is also a major concern, calling into question the very reliability of online information and exacerbating fears around electoral manipulation.

Faced with these challenges, adequate regulation is becoming essential in the eyes of a large majority of users. Yet, only two in five individuals consider current regulatory frameworks sufficient. There is thus a strong expectation for concerted international regulation, combined with effective national governance involving both governments and industry. However, lack of awareness of existing laws and regulations on AI remains a worrying reality, potentially limiting their effectiveness.

In this context, organizational assurance mechanisms play a crucial role in building lasting trust in AI. Users clearly express that they would be more inclined to trust AI systems if safeguards such as liability policies, human oversight, or independent monitoring mechanisms were in place.

Finally, special attention must be paid to AI education and training, which are still largely insufficient. Increased investment in digital literacy and education on the proper use of AI tools could help alleviate this ambivalence and strengthen responsible and critical adoption of these technologies.

The future of AI can only be safely embraced through governance that places humans at the center, incorporating education, balanced regulation, and organizational transparency as fundamental pillars. Only then can the considerable benefits of AI be fully realized while minimizing its risks.

To go further: this reflection on responsible adoption continues with Our analysis of Shadow AI, faster than governance et our guide to creating an AI usage charter.