TRANTOR LABS has formally released the current finalized version of its first general paper on civilizational risk:
The Risk-Bearing Subject of AI-Related Existential Risk to Humanity: From Species Survival to the Realization of Subjecthood under Civilizational Conditions
The first public version of this paper was released on June 7, 2026. Over the following two weeks, the paper went through three public version updates. Version 3, released on June 17, 2026, is the current finalized version of this paper.
This release marks an important step in the public development of TRANTOR LABS’ research program. It is not merely the publication of a single paper. It is the first formal research anchor in our broader work on human existential risk in the age of AGI.
The paper can be accessed and downloaded here: Zenodo: https://zenodo.org/records/20728025; PhilPapers: https://philpapers.org/rec/YUATRS.
Why this paper comes first
Before identifying AI-related existential risks, we must first clarify a more basic question: who, or what, is the subject that bears the risk?
In many discussions of AI safety, AI governance, and AI risk, people say that AI may threaten “humanity.” But the term “humanity” often points to different things in different contexts. Sometimes it refers to human preferences. Sometimes it refers to human lives. Sometimes it refers to social order, public institutions, species continuity, or human values.
If the risk-bearing subject is not clarified, risk analysis can slide across different dimensions. A discussion may begin with model behavior, move to biological survival, then shift to institutional stability, and then to values, agency, or meaning, without clearly explaining how these dimensions are related.
This paper therefore begins with the most basic question: when we say that AI may pose an existential risk to humanity, who is actually at risk?
Central thesis
The paper argues that the risk-bearing subject of AI-related human existential risk remains humanity as an embodied species.
It does not add civilization, institutions, culture, or the philosophical kernel as new risk-bearing subjects. What it expands is not the risk-bearing subject, but the ways in which one and the same human subject may suffer existential harm.
Humanity may be harmed at the biological-survival layer, for example through extinction, permanent incapacitation, or the irreversible closure of the human future. But humanity may also be harmed at the civilizational-conditions layer: the layer of conditions through which humans realize subjecthood through language, knowledge, institutions, responsibility, judgment, public reason, education, memory, and meaning.
In this sense, human existential risk in the age of AGI should not be understood only as the biological destruction of Homo sapiens. It may also include structural damage to the civilizational conditions under which humans continue as agents of judgment, responsibility, reality-seeking, collective governance, and meaning-making.
Key concepts introduced in the paper
This paper introduces several concepts that will become foundational for the research that follows.
The first is “humanity as an embodied species.” The paper insists that the risk-bearing subject remains humanity. Humanity here is not a pure information structure, a preference aggregate, or an abstract rational agent, but embodied human beings who live, act, suffer, judge, inherit, and bear responsibility within time.
The paper then distinguishes between the biological-survival layer and the civilizational-conditions layer. The former concerns the survival of humanity as a species; the latter concerns the civilizational conditions through which humans realize subjecthood. Civilizational conditions refer to the shared structures that allow humans to know reality, form judgment, assign responsibility, participate in institutions, inherit meaning, and act toward the future.
Within civilizational conditions, the paper further identifies the “philosophical kernel.” This does not refer to academic philosophy as a discipline, nor to any fixed theoretical doctrine. It refers to the category-calibration mechanism within civilizational conditions that keeps foundational categories such as truth, evidence, responsibility, subjecthood, value, and meaning operational.
The paper also introduces two mechanisms through which AI may damage the realization of human subjecthood: substitutive deprivation and mediated reconfiguration. The former refers to AI replacing humans in substantive participation within functional systems of civilization. The latter refers to AI reorganizing the conditions under which human judgment, responsibility, and meaning are formed, while still preserving humans in formal positions.
Finally, the paper proposes four thresholds for determining when damage to civilizational conditions enters the scope of existential risk. Damage at the civilizational-conditions layer becomes existentially relevant only when it acts upon foundational civilizational conditions, has the capacity to diffuse across domains and civilizational interfaces, may cause irreversible or difficult-to-reverse structural harm, and ultimately converges into overall, intergenerational damage to humanity’s capacity as a future-oriented subject.
What this paper does not claim
This paper is not a complete taxonomy of AI risks, nor is it a governance proposal. It does not claim that every social harm caused by AI should be treated as an existential risk, and it does not reduce the importance of biological-survival risk or species-extinction risk.
It also does not claim that civilization, institutions, or philosophical concepts themselves become independent risk-bearing subjects. Its central claim is more precise: one and the same human subject may suffer existential harm not only through biological destruction, but also through damage to the civilizational conditions under which human subjecthood continues to be realized.
This boundary is important. It allows the paper to preserve the baseline importance of species survival in existing existential-risk discussions, while also giving conceptual space to slower, more hidden, and more structural forms of damage to civilizational conditions in the age of AI.
Why this paper matters for the research program
This paper establishes the first layer of TRANTOR LABS’ research program on human existential risk in the age of AGI.
It answers the first question: who is at risk?
The following general papers on civilizational risk will build on this foundation and ask: what are the primary risks? How do these risks emerge? How do they affect the conditions under which humans act, know, govern, judge, and make meaning?
Without this first step, later discussions of survival risk, reality risk, institutional risk, agency risk, and meaning risk would lack a stable ontological foundation. This is why the first paper does not begin with a risk list, but with the risk-bearing subject itself.
What comes next
Around this paper, TRANTOR LABS will continue to publish explanatory essays, research notes, and reader-facing theoretical teardowns. These will include essays on why AI governance must first clarify what “humanity” means, why species survival is not the same as civilizational survival, why formal participation is not substantive participation, why the philosophical kernel matters for AI risk, and when damage to civilizational conditions becomes existential risk.
The next stage of the research will move from “who is at risk” to “what are the primary risks.”
This paper is therefore both a beginning and a foundation. It marks the point at which TRANTOR LABS’ internally consolidated research begins to formally enter public scholarly space.
