Artificial intelligence is becoming an increasingly pervasive presence, both in everyday software applications and at the most advanced frontiers of human-machine interaction, raising crucial questions about control and ethical AI.
What happened
The integration of artificial intelligence into our daily digital lives continues to grow, bringing new challenges in terms of user control and autonomy. A recent example is the introduction of AI functionalities in popular tools like Google Docs, where users are confronted with proactive suggestions generated by Gemini, Google's AI. While these features can boost productivity, they also raise the question of the need to be able to disable such assistants to maintain full command of one's digital workspace, as highlighted by TechCrunch AI. The demand for an "off switch" for AI reflects a fundamental desire for agency from users, who do not want AI imposed, but rather integrated in a controlled and optional manner.
In parallel, the frontier of human-machine interaction extends far beyond software, reaching directly into the human brain. Trials on brain-computer interfaces (BCI) are taking off, with significant progress enabling people with severe disabilities to communicate and interact with the digital world. The case of Casey Harrell, a man with ALS who has used a brain implant for almost three years to communicate, demonstrates the transformative potential of these technologies MIT Technology Review AI. These devices, such as those developed by companies like Neuralink or Synchron, promise to restore autonomy and quality of life, but also open a complex ethical debate about human augmentation, the privacy of neural data, and potential inequality in access to such technologies.
Why it matters
The ubiquity of AI in software and the emergence of BCIs converge on a central point: the profound impact of artificial intelligence on the individual. The ability to disable AI features in a text editor is not just a matter of preference, but of preserving cognitive autonomy and freedom of choice. If AI becomes too intrusive, it can alter human decision-making processes, creativity, and even self-perception. On the other hand, BCIs promise a radical improvement in quality of life for those suffering from severe conditions, but at the same time, they push us to reflect on the boundaries between repair and enhancement, and who will have access to these innovations.
A crucial aspect in this scenario is trust in artificial intelligence systems. The reliability and impartiality of the metrics used to evaluate AI performance are fundamental, yet often underestimated. As highlighted by MIT Technology Review AI, metrics, while useful, can obscure or corrupt the understanding of reality, leading to erroneous decisions or biases. This is particularly true for AI, where complex algorithms can generate results that are difficult to interpret or critically judge without a clear understanding of the underlying metrics and their limitations. Uncritical reliance on AI metrics can lead to unfair outcomes or a false perception of objectivity.
The HDAI perspective
The increasing integration of AI into our daily lives and our very biology highlights an urgent need: for AI governance that is genuinely Human Driven AI. It is not just about regulating technology, but about ensuring that the development and implementation of AI are guided by solid ethical principles, placing human well-being, autonomy, and dignity at the center. The ability to disable AI in software is a small, but significant, act of control that users should always have. Similarly, BCIs, while revolutionary, require public debate and clear regulations to prevent abuse, ensure equitable access, and protect neural privacy.
The real challenge is to build artificial intelligence systems that are not only powerful but also understandable, controllable, and aligned with human values. This approach will be central to discussions at the HDAI Summit 2026, where experts from around the world will discuss how Italy and Europe can lead the development of artificial intelligence that is a driver of ethical and inclusive progress. It is essential that AI evaluation is not based solely on efficiency metrics but also includes social and ethical impact indicators, promoting a culture of responsibility and transparency.
What to watch
In the coming years, the evolution of BCIs and the increasingly deep integration of AI into common devices will require constant monitoring. It will be crucial to observe how regulations, such as the EU AI Act, will adapt to these new technological frontiers and whether they will succeed in balancing innovation and the protection of individual rights. Attention will also shift to user education, so that they can fully understand the implications of AI and make informed choices regarding its use.