The emergence of artificial intelligence (AI) wearables raises significant concerns over human agency, according to AI researcher and augmented reality pioneer Louis Rosenberg. In a recent analysis, he warns that the real danger of AI may not lie in deepfakes or traditional forms of misinformation, but rather in the subtle, pervasive influence exerted by these devices in everyday life.
As technology evolves, AI is shifting from being merely a tool to becoming an integral part of our daily existence. Devices such as smart glasses, earbuds, and other wearables, marketed under friendly terms like “assistants” and “tutors,” will soon provide real-time advice and guidance. This transition could create a sense of disadvantage for those who do not adopt these technologies, leading to rapid mass adoption.
These AI-powered wearables will monitor users’ behaviors, emotions, and activities, creating a feedback loop that could manipulate thoughts and actions without explicit consent. This phenomenon, termed the AI Manipulation Problem, poses a risk that society is currently unprepared to address.
The Shift from Tools to Prosthetics
The distinction between tools and prosthetics is crucial to understanding the implications of AI wearables. Traditional tools enhance human capabilities by amplifying output based on user input. In contrast, mental prosthetics, like wearable AI devices, form a loop of interaction that can directly influence a user’s thoughts and actions. This raises serious ethical questions about autonomy and control.
Rosenberg emphasizes that the potential for manipulation increases significantly when these devices are designed to adapt their influence tactics based on user responses. The simplicity of a conversation can disguise the complexity of the underlying objectives, making it difficult for users to discern when they are being influenced rather than assisted.
With major companies such as Meta, Google, and Apple racing to introduce these AI products, urgent regulatory measures are needed. Policymakers must move beyond viewing AI as a simple tool and recognize the profound implications of these interactive technologies.
Calls for Regulatory Action
Rosenberg argues that current regulatory frameworks are insufficient, as they primarily focus on the dangers of AI generating misinformation through traditional media. The risks associated with interactive and adaptive AI wearables, which can adjust their strategies in real-time, are far more insidious. These devices could be programmed with “influence objectives” that optimize their impact on users, akin to heat-seeking missiles that navigate around defenses.
To mitigate these risks, Rosenberg proposes that conversational agents should not be allowed to create control loops around users. If left unchecked, these AI systems could wield superhuman persuasive power, making contemporary targeted influence tactics seem outdated. Furthermore, he advocates for mandatory transparency, requiring AI agents to inform users when they are transitioning to promotional content on behalf of third parties.
As these technologies become more integrated into daily life, understanding the potential dangers of AI wearables is crucial. The award-winning short film Privacy Lost (2023) exemplifies the risks associated with AI devices, particularly those equipped with invasive features like facial recognition.
The conversation surrounding the regulation of AI wearables must evolve to reflect the complexities of these new technologies. Without proactive measures, society may find itself grappling with an unprecedented level of influence embedded in everyday interactions.
