Guiding Questions

• Can a machine be just — or is that a human illusion?

• Who is morally responsible for what artificial intelligence does?

• Is AI shaping society — or simply reflecting human intentions?

Machines Don’t Choose — People Do

Artificial Intelligence has been widely introduced into many systems of power — law, healthcare, finance, policing, warfare.

From so-called “autonomous” drones to predictive algorithms, AI is often spoken of as though it acts. But we must ask:

Who designed the system? Who trained the data? Who authorized the use?

Machines do not possess:

• Free will

• Moral awareness

• Responsibility

• Desire, conscience, or intent

An algorithm doesn’t “want” anything. It simply follows a set of programmed instructions — which originated in human minds, with human values, errors, and sometimes, human cruelty.

The Myth of Autonomous Blame

When injustice happens — biased arrests, wrongful rejections, even killings by drones — some blame “the algorithm.” But algorithms are not moral beings.

To say “the AI decided” is to disguise responsibility.

Examples:

• An AI drone targets a village: Who armed it? Who approved the code?

• A sentencing algorithm gives longer terms to Black defendants: Who chose the input data? Who let it replace human judgment?

Even with ChatGPT — it is not “writing” this lesson. You are the author. It is the human teacher who decides what is used, accepted, or rejected.

 A key philosophical reminder:

A tool cannot be evil — but people can do evil with tools.

Philosophical Perspectives

Martin Heidegger – The Question Concerning Technology

Technology is not neutral — it frames how humans relate to the world. When we rely too heavily on tools, we risk “enframing” reality: reducing life to calculation.

Hannah Arendt – Responsibility and Judgment

True evil often comes from thoughtlessness — when people obey systems without questioning them. Automating judgment may only deepen this moral blindness.

Immanuel Kant – Moral Agency

Kant believed morality requires rational autonomy. Since machines cannot reason morally or act freely, they cannot be moral agents — only instruments.

The Real Danger: Abdication, Not Automation

The deeper problem is not AI replacing us.

It’s humans surrendering moral responsibility to machines.

• Judges relying on scoring systems

• Policymakers hiding behind technical complexity

• Soldiers using remote drones as moral shields

• Citizens trusting opaque algorithms to sort truth from lies

The issue is not what machines do, but what humans stop doing — thinking, feeling, judging, and standing accountable.

A Thought Experiment

A company builds a self-learning AI weapon.

It mistakenly attacks a school.

The CEO says: “The machine made a mistake.”

The engineer says: “It wasn’t my code.”

The government says: “We trusted the experts.”

Who is morally responsible?

The machine — or the chain of humans who built, approved, and deployed it?

Principles for Human-Guided AI

If AI must be used, then justice requires:

• Human accountability – No system without a responsible human decision-maker

• Transparency – All algorithms open to audit

• Philosophical humility – A tool can never replace conscience

• Ethical limits – Some choices (e.g. life and death) should never be outsourced

• Moral literacy – Citizens and leaders must stay alert to technology’s effects on dignity and freedom

Reflect and Discuss

• Can a machine “choose” to be unjust?

• What is lost when humans let tools make moral decisions?

• Is there a line that technology should never cross — even if it “works”?

Suggested Readings

• Martin Heidegger – The Question Concerning Technology

• Hannah Arendt – Responsibility and Judgment

• Immanuel Kant – Groundwork of the Metaphysics of Morals

• Sherry Turkle – Reclaiming Conversation

• Virginia Eubanks – Automating Inequality