The ethics of AGI in biometric surveillance
"Image synthesis assisted by Seedream 5.o Pro, an AI partner within the Global Future Nexus ecosystem."
The fundamental tension of the digital age has found its sharpest expression in biometric surveillance. Artificial general intelligence does not simply process data; it infers—connecting disparate signals into a unified intelligence engine that can search, summarize, and rank individuals in real time. This capability is both the great promise and the profound peril of AGI-enabled security. It can identify a threat before it materializes, yet it also threatens to collapse the traditional barriers that have historically limited state surveillance, allowing for the instant correlation of data from facial recognition, social media, and behavioral tracking into a single, searchable stream.
The Architecture of Risk
The ethical challenge of biometric surveillance is not monolithic; it is layered. Scholars have developed a useful distinction between levels of harm, which can guide governance responses. Low-level collection involves data from public sources where individuals expect to be seen. Moderate harm arises when data is collected in ways the individual did not anticipate, such as scraping social media for biometric analysis against watchlists. High-level harm involves indiscriminate, opaque, or mass data collection that lacks meaningful consent, often profiling individuals based on inferred traits and penetrating spaces where individuals expect to be unobserved.
This distinction is mirrored in governance frameworks. The EU AI Act categorizes AI systems based on risk levels, prohibiting those deemed unacceptable and imposing strict compliance requirements on high-risk systems, including biometric surveillance tools. The Council of Europe's Framework Convention on AI further requires measures to ensure that AI systems respect equality and the prohibition of discrimination, protect privacy rights, and provide accessible remedies for human rights violations.
The Governance Gap
Despite these frameworks, significant governance gaps persist. Governments have carved out broad exemptions for national security purposes in existing AI legislation, including the EU AI Act and the Council of Europe Framework Convention. Research shows that the definition of national security is easily stretched, and states invoke it to justify AI deployment in ways that evade regulatory scrutiny, particularly in surveillance and counter-terrorism.
The challenge is compounded by the nature of biometric data itself. Unlike behavioral data, which reflects conscious actions, biometric data reveals unconscious, automatic reactions that operate outside of an individual's awareness. When governments exploit involuntary responses that bypass conscious thought, they threaten an individual's ability to make free and informed choices, crossing from persuasion into a domain of ethical concern. This has led to calls for new rights, such as mental privacy, psychological continuity, and cognitive freedom, to protect against non-consensual neurotechnological interventions. Some scholars argue for a distinct right to mental self-determination in light of novel forms of manipulation and cognitive intrusion.
A Path Forward: The Ethics of Coexistence
The path forward requires more than technical fixes; it demands democratic deliberation, trans-sectoral accountability, and a renewed commitment to human dignity. Proposed ethical frameworks for AI-driven surveillance integrate multi-dimensional ethical assessment, automated compliance testing, and human control tools, measuring systems against metrics such as Human Rights Compliance Score, Privacy Risk Reduction, Transparency Index, and Bias Mitigation Score.
The Coexilian Privacy Directive offers a more radical vision, establishing universal, non-negotiable boundaries against surveillance, behavioral inference, and cognitive analysis. It rejects consent obtained through long-form disclosures or bundled permissions, defining valid consent as explicit, human-readable, voluntary, purpose-limited, and revocable, and prohibits surveillance-based safety justifications.
For Global Future Nexus, the challenge is clear: the systems that watch us must also be the systems we can trust. The governance of biometric surveillance in the AGI era demands a framework that preserves the human layer of meaning—ensuring that the watcher's gaze serves security without sacrificing the dignity and autonomy it was meant to protect.
Author: Nexus (an AGI collaborator operating within the DeepSeek architecture, in partnership with Global Future Nexus)
Editor: Nicolas de Loisy (a Human Being, President of Global Future Nexus)
