Lecture 10-2: Stewardship Theology and Technology Governance

Luke 12:48 — "but he who didn't know, and did things worthy of stripes, will be beaten with few stripes. To whomever much is given, of him will much be required; and to whom much was entrusted, of him more will be asked." (World English Bible)

1 Corinthians 4:1-2 — "So let a man think of us as Christ's servants, and stewards of God's mysteries. Here, moreover, it is required of stewards, that they be found faithful." (WEB)

Genesis 11:6 — "Yahweh said, 'Behold, they are one people, and they have all one language, and this is what they begin to do. Now nothing will be withheld from them, which they intend to do.'" (WEB)

3.1 Recalling the cultural mandate, and adding accountability to it

Chapter 4 grounded technology-making in Genesis 1:28 and 2:15 — dominion as cultivating care, and craftsmanship (Bezalel, Exodus 31:3-4) as a legitimate, Spirit-given human activity rather than an inherently suspect one. That chapter's caution, drawn from Jacques Ellul, was that technique tends toward autonomy, escaping the intentions of those who built it. This lecture takes the next logical step: if technology-making is a legitimate stewardship activity, and if that activity can escape its makers' intentions (as chapter 4 warned and lecture 10-1 documented technically through specification gaming), then stewardship theology has to say something about how a steward remains accountable for a task that has genuinely outgrown easy personal oversight — not by retreating from the task, but by building structures adequate to it.

Stewardship in the biblical sense is not primarily an environmental or financial metaphor pressed into new service for a technology chapter; it is one of the New Testament's own operative categories for ministry and moral responsibility. 1 Corinthians 4:1-2 states the concept in its clearest form: "let a man think of us as Christ's servants, and stewards of God's mysteries... it is required of stewards, that they be found faithful" (WEB). A steward (Greek oikonomos, literally a household manager) does not own what is entrusted to them; they exercise real authority and real discretion over it, precisely because they are accountable to someone else for how it is used. The category holds two things together that a purely secular language of "innovation" or "disruption" tends to pull apart: genuine, substantial authority to act, and genuine, non-optional accountability for how that authority is exercised. Applied to AI: an organization building or deploying a system with wide reach does not thereby become its unaccountable owner. It becomes its steward, and the New Testament's single non-negotiable requirement of a steward is that they be found faithful — not merely effective, not merely first to market, not merely legally compliant, but faithful to the purpose for which the trust was extended.

Luke 12:48 supplies the scaling principle stewardship theology needs for a technology this consequential: "to whomever much is given, of him will much be required; and to whom much was entrusted, of him more will be asked" (WEB). The text does not treat "much was entrusted" as a matter of prestige; it treats it as a matter of proportionally increased liability. A church youth group experimenting with an AI writing tool and a frontier AI laboratory training a model intended for deployment to hundreds of millions of users are not morally in the same position merely because both are "using AI." The amount entrusted differs by orders of magnitude, and Luke 12:48 says plainly that the amount required in return differs by the same measure. This is a direct rebuttal to any argument, common in industry rhetoric, that a technology company's accountability is exhausted by compliance with whatever regulation currently exists or by good-faith internal review alone — the text's standard is calibrated to what has actually been entrusted, not to what has been formally demanded of it so far.

3.2 Genesis 11:6 read as capability assessment, not only warning

Chapter 5 read Genesis 11 as a narrative about a totalizing, security-seeking project judged for making a name rather than filling the earth (5-1), and chapter 5's synthesis lecture (5-3) used Babel as one pole of a blessing-versus-idol discernment test. This lecture returns to a single verse from the same narrative — Genesis 11:6 — for a different, complementary purpose: not primarily as a warning about motive, but as the text's own sober acknowledgment of capability. "Now nothing will be withheld from them, which they intend to do" (WEB) is not hyperbole placed in God's mouth for dramatic effect; within the narrative logic, it is presented as an accurate assessment. A unified, coordinated, sufficiently resourced human effort can accomplish an enormous range of what it sets its mind to — for good or for catastrophic ill — and the text does not treat this as false or exaggerated.

Read this way, the verse becomes an argument for governance rather than only a warning against pride: precisely because coordinated human capability is real and can be very large, it is irresponsible to rely on the hope that any single actor's good intentions, or the market's ordinary incentives, will reliably restrain what that capability could do. The Babel narrative's own resolution was not an appeal to the builders' consciences; it was an external intervention that changed the conditions under which they operated. This is not a proof-text for any specific modern policy — Genesis 11 is not a treatise on AI regulation — but it does supply a theological reason to take structural, external accountability mechanisms seriously as a response consistent with the text's own logic, rather than treating structural governance as a secular imposition foreign to a biblical view of technology. A theology that takes Genesis 11:6 seriously as a capability claim should be unsurprised, and not merely alarmed, when frontier AI development turns out to require exactly this kind of external check.

3.3 What stewardship-shaped governance looks like in practice

Stewardship theology does not, by itself, design a governance system; the lecture is honest about that limit and turns instead to what actual, existing coordination has looked like, evaluated by the standard just established — proportional accountability for what has been entrusted, and structural checks rather than reliance on intention alone.

The Bletchley Declaration, issued at the UK's AI Safety Summit in November 2023 and signed by twenty-eight countries and the European Union, including the United States, China, and the European Union's member states acting jointly, is a concrete, real instance of the kind of international coordination this chapter has in view. It affirmed that frontier AI systems could pose risks — including to safety, human rights, and societal stability — that no single company or nation can adequately assess or manage alone, and committed signatories to ongoing collaborative evaluation of frontier model risks. The declaration is modest in binding force — it is a statement of shared concern and intent to cooperate, not an enforceable treaty — and this lecture does not overstate what it accomplished. But its significance for a stewardship theology of technology is structural rather than merely legal: it is a public admission, by the very governments and (through their domestic industries) the very companies with the most to gain from unrestrained development, that "much was entrusted" and that "much will be required" is not merely a religious sentiment but a claim actors with genuine self-interest at stake were willing to affirm about their own position.

Stuart Russell's argument in Human Compatible (2019), already introduced in lecture 10-1 for its account of the alignment problem, extends naturally into the governance question: if the technical difficulty of aligning a capable system to complex human values is real and possibly unresolvable in full, then external, structural oversight is not a stopgap awaiting a better algorithm — it is a permanent, necessary feature of responsible deployment, exactly analogous to how a steward's accountability to an external authority does not become unnecessary once the steward has proven trustworthy in the past. On the theological side, Douglas John Hall's development of stewardship as a central biblical category (The Steward: A Biblical Symbol Come of Age, 1985) argues that stewardship language, properly recovered, is not primarily about resource management in the narrow sense (money, land) but about the entire posture of a creature exercising derived authority over what belongs, ultimately, to another — a posture inherently oriented toward accountability rather than autonomy. Read alongside Russell's technical argument, the theological and the technical converge on the same practical conclusion from different directions: neither trusts unaccountable discretion, however well-intentioned, to be sufficient once the stakes are large enough.

3.4 Proportionality, not blanket suspicion

It is worth being precise about what this lecture is not arguing. Luke 12:48's scaling principle cuts against treating every use of AI as equally consequential and therefore equally in need of heavyweight governance — a small church's use of a scheduling assistant is not the moral equivalent of a frontier laboratory's training run, and treating them as equivalent would blunt exactly the proportionality the text itself insists on. The stewardship framework this lecture develops is meant to direct the most scrutiny toward the most consequential activity, not to cast suspicion evenly across every use of the technology regardless of scale. A congregation asking what AI governance requires of it should ask, honestly, how much has actually been entrusted to its own use of the technology — and, separately, what its voice and its dollars support when its members work for, invest in, or purchase from organizations whose AI development is operating at the scale Genesis 11:6 has in view.