[D66] On Demis Hassabis’s Neutrality Claim, and Why It Fails

[René Oudeweg]

On Demis Hassabis’s Neutrality Claim, and Why It Fails

When Demis Hassabis says that “technology is neutral,” he is not making 
an innocent philosophical claim. He is performing an act of power 
laundering.

This line does not emerge from confusion. It emerges from position. It 
is what someone says when they sit at the helm of one of the most 
consequential technological enterprises in human history and wish to 
retain the authority of that position without the moral exposure it 
entails. “Neutrality” becomes a shield: smooth, abstract, and 
conveniently impermeable to blame.

Let’s be clear about what is happening here. Hassabis is not a detached 
observer commenting on tools made by others. He is a central architect 
of systems that concentrate knowledge, labor, capital, and 
decision-making capacity at unprecedented scale. To claim neutrality 
from that vantage point is not modesty—it is audacity.

Technology does not drift into existence. It is willed. It is directed. 
It is sculpted by people who decide what gets funded, what gets scaled, 
what gets scrapped, and what gets shoved into the world before society 
has any meaningful say. When Hassabis speaks of neutrality, he is 
erasing those decisions with a wave of the hand, pretending that intent 
dissolves the moment it is expressed in code.

This is not a harmless abstraction. It is a narrative designed to 
dissolve accountability.

If technology is neutral, then no one is responsible for the labor 
markets it destabilizes. No one is responsible for the creative fields 
it strip-mines. No one is responsible for the epistemic chaos it 
accelerates, the surveillance infrastructures it feeds, or the power 
asymmetries it entrenches. These become “societal questions,” 
conveniently outsourced to a future that will never quite catch up.

Neutrality is not a description here—it is an excuse.

What makes this posture especially galling is that AI, of all fields, 
exposes the lie most clearly. AI systems are not neutral even in theory. 
They are trained on human data, optimized toward specific objectives, 
constrained by institutional priorities, and deployed in environments 
shaped by inequality. Every layer is soaked in value judgments. To deny 
this is to insult the intelligence of anyone who actually understands 
how these systems work.

So when Hassabis invokes neutrality, he is not speaking as a scientist 
clarifying a misunderstanding. He is speaking as a CEO protecting a 
perimeter.

There is also a deeper arrogance embedded in this claim: the assumption 
that ethical concern is something that happens after the breakthrough, 
that morality is downstream of brilliance. First we build godlike tools, 
then—if time permits—we ask how they affect everyone else. This is the 
worldview of someone who has never been on the wrong side of 
technological “progress,” someone for whom disruption is an intellectual 
puzzle rather than a lived catastrophe.

“Technology is neutral” is what you say when you want to be seen as 
building the future while remaining curiously absent from its wreckage.

It is also a way of delegitimizing dissent. If the technology is 
neutral, then critics are irrational. Workers resisting automation are 
sentimental. Artists objecting to mass extraction are Luddites. 
Communities fearing surveillance are paranoid. Neutrality becomes a 
cudgel: a way to frame opposition as ignorance rather than as moral 
disagreement.

And let’s dispense with the pretense that this is an abstract debate. 
Statements like this shape policy, regulation, and public tolerance. 
They normalize the idea that those who build the most powerful systems 
should also be the least accountable for their outcomes. That is not 
accidental. It is extremely useful.

The irony is that Hassabis does not need this lie. If anything, his 
influence demands the opposite: explicit ownership of values, 
trade-offs, and consequences. But that would require admitting that AI 
development is not just an engineering problem—it is an exercise in 
governance. And governance invites scrutiny.

So neutrality is offered instead. Clean. Bloodless. Convenient.

But history is not fooled by this move. No transformative technology has 
ever been neutral in effect, and no one at the center of its creation 
has ever been merely a bystander. Those who claim otherwise are not 
avoiding politics; they are practicing it quietly, while insisting they 
are above it.

Demis Hassabis can continue to say that technology is neutral. But what 
that really communicates is this: I want the power to change the world 
without being held to account for the world that results.

That is not a philosophical position.
It is a refusal.