Digital platforms have become the engine of current commercial exchanges. But AI, a central component of these systems, strongly impacts the fundamental notion of trust within digital ecosystems.
Published on 2024-12-05 by Nathalie Lamborghini Dumas.
Digital platforms have become the engine of current commercial exchanges, connecting companies worldwide in a few clicks. They simplify supply chains, provide instant access to strategic data and adapt operations in real time. But this connectivity comes with new risks: artificial intelligence (AI), a central component of these systems, strongly impacts the fundamental notion of trust within digital ecosystems.
According to a global survey conducted by Deloitte in 2024, nearly 50% of boards of directors have not yet integrated AI into their strategic agenda.
Yet, the role of AI in business transformation, and particularly in platform businesses, continues to grow, generating opportunities but also major challenges related to governance, ethical risks and transparency.
#### 1. AI: between innovation catalyst and critical governance challenge
Artificial intelligence boosts digital platforms, but it also amplifies already existing risks while introducing new challenges.
Among amplified risks, data protection remains crucial. Platforms that collect and analyze massive volumes of data expose their ecosystems to leaks or abuses. A Deloitte survey reveals that 72% of employees use unapproved tools, accentuating these risks. Algorithmic transparency is also undermined, as Amazon demonstrated in 2017 with a discriminatory recruiting tool, highlighting the danger of hidden biases in AI systems. Finally, regulatory compliance questions become more complex, demanding increased vigilance from companies.
Meanwhile, the emergence of generative AI brings unprecedented challenges. Deepfakes and algorithmic hallucinations pose serious misinformation problems, undermining the credibility of platforms and their stakeholders. Copyright and intellectual property represent another tension zone, with disputes over the use of protected content to train these models. Cybersecurity is also a major challenge, as generative AI systems expand the attack surface for malicious actors. Finally, the lack of skills within organizations hampers their ability to anticipate and manage these challenges, even at the highest level: 79% of boards acknowledge limited understanding of AI, according to Deloitte.
To meet these challenges, companies must adopt proactive governance, strengthen transparency and invest in training. Without this, AI risks compromising trust in digital ecosystems, transforming a strategic opportunity into a source of vulnerabilities.
#### 2. Building trustworthy AI: winning strategies
To strengthen trust in AI, platforms must adopt concrete measures:
AI governance cannot be left to the discretion of technical teams. It must be led by governing and controlling bodies. Companies must: - Put AI on the strategic agenda of boards of directors, regularly discussing associated risks and opportunities. - Create specialized committees, as Deloitte proposes, to manage ethical and regulatory risks related to AI. - Establish ethical principles for AI deployment.
Algorithms must be understandable, even for non-technicians. Spotify, for example, explains why a song is recommended to its users, thus strengthening their trust. B2B platforms can draw inspiration from this approach by explaining the selection criteria of their commercial recommendations.
Educating users and management teams is essential. According to Deloitte, 40% of companies organize AI-dedicated training for their boards of directors. Better understanding of AI bridges the gap between technological opportunities and risk management.
Tools like federated learning, used by Google in its Gboard keyboard, allow training algorithms locally without transferring sensitive data. These approaches minimize leak risks while ensuring optimal performance.
#### 3. Example: Maersk and AI for global logistics management
Maersk, the maritime transport leader, introduced Captain Peter, a virtual assistant powered by artificial intelligence, to revolutionize supply chain management. This system offers total transparency by tracking refrigerated containers in real time and proactively communicating with clients via their preferred channels (SMS, email). Clients receive detailed updates on temperature, atmosphere and itinerary, strengthening their trust through increased visibility over their shipments.
AI also allows Captain Peter to analyze real-time data to anticipate problems, such as delays or anomalies in transport conditions. For example, it can alert clients in advance so they act quickly, thus avoiding costly losses. This operational reliability improves efficiency and reinforces the perception of a reliable service.
Finally, Maersk uses AI to personalize the information provided to each client's specific needs, while simplifying processes through automation. This personalization and simplification reinforce customer satisfaction, while preparing the future with continuous improvement capabilities, such as cargo damage prediction.
Challenges: to maintain this trust, Maersk must guarantee rigorous protection of sensitive data, ensure total transparency on algorithmic decisions and address ethical questions related to equity and AI's impact on employment.
Lessons learned: through proactive governance and strategic use of AI, Maersk has transformed a complex industry into a transparent, personalized and reliable customer experience, while managing risks to maintain lasting trust.
#### Conclusion: proactive governance for trustworthy AI
Artificial intelligence boosts digital platforms, but this transformation cannot succeed without rigorous governance. Platforms must balance automation with human judgment to avoid drifts linked to blind dependence on algorithms. Transparency is essential: decisions made by AI must be understandable, both for users and partners.
Boards of directors must integrate AI into their strategy and create dedicated frameworks to supervise its development. Specialized committees can manage ethical challenges, operational risks and regulatory compliance. In parallel, education of management teams and users is crucial to strengthen understanding of AI tools and maximize their impact while limiting risks.
Adopting these measures allows platforms to transform challenges (such as biases, misinformation or privacy violations) into opportunities to build lasting relationships with their users. Ultimately, AI is not just an innovation lever, but also a strategic tool for strengthening trust in increasingly complex digital ecosystems.
As one expert interviewed by Deloitte summarized: "the greatest risk of AI is not taking any."