Trustworthy AI in the Age of Pervasive Computing and Big Data

Data collection in pervasive computing creates dual-use risks: The paper identifies that while continuous monitoring through connected devices can improve service quality, the same data infrastructure enables surveillance and manipulative advertising—highlighting that trustworthy AI requires careful governance of data collection practices, not just how data is used.

AI in high-stakes domains demands ethical integration from the start: The paper emphasizes that AI systems applied to sensitive fields like healthcare, justice, and human resources raise trustworthiness concerns that cannot be addressed retroactively, suggesting that ethical considerations must be built into design and development processes rather than treated as an afterthought.

Trust in AI is fundamentally an ethics problem: The paper formalizes trustworthy AI through three interconnected ethical dimensions—the ethics of algorithms, the ethics of data, and the ethics of practice—indicating that technical solutions alone are insufficient without addressing the broader ethical frameworks governing AI systems.

Practical implementation is possible: The paper demonstrates that trustworthy AI requirements can be concretely integrated into system design and development, using a smart cities use-case as evidence that ethical principles are not merely theoretical but can guide real-world AI deployment.