The Future(s) of HumA(I)nity: Human Agency, Agentic AI and Agentified AI Governance
JMCE JUST-AI
Info
JUST-AI JEAN MONNET CENTRE OF EXCELLENCE
CLOSING CONFERENCE
Within this conceptual framework, AI systems may be understood as mediating freedom in distinct and normatively significant ways. Rather than conceptualising such systems as either neutral tools or autonomous agents, we propose to examine how they participate in shaping the conditions under which freedom is exercised. In this regard, a useful distinction can be drawn between freedom-supporting and freedom-reducing forms of mediation. The former includes practices that enhance users’ capacity for reflection, articulation, and informed decision-making, such as clarifying arguments, expanding interpretive options, and improving conceptual or textual expression within user-defined goals. The latter refers to forms of influence that constrain, distort, or bypass deliberation, thereby undermining the user’s capacity for self-determination.
Crucially, both modalities involve some form of influence over action; the normative difference does not lie in the presence or absence of mediation, but in the conditions under which such mediation operates. These conditions may be articulated through three interrelated criteria: transparency, understood as the possibility for users to recognise and reflect upon the system’s role in shaping outcomes; revisability, understood as the user’s capacity to contest, modify, or override system outputs; and alignment, understood as the degree to which system behaviour remains responsive to the user’s own ends rather than substituting them. Taken together, these criteria offer a way of operationalising the concept of freedom in AI-mediated environments without reducing it either to non-interference or to an unrestricted notion of technological co-production.
The JUST-AI Jean Monnet Centre of Excellence closing conference asks a fundamental question: whose agency should align with and deliver fairness? Is it a human actor - and if so, which one? Is it an artificial, agency-capable artefact or software system? Or is it an institutional actor, such as a public or private agency tasked with monitoring and governing human-AI interactions?
Building on fairness as the central theme that has connected previous editions of the JUST-AI Jean Monnet Summer Colloquia, the closing conference of the JUST-AI Jean Monnet Centre of Excellence invites scholars at all stages of their careers to revisit the question of humancentrism.
