Question: If AI recommended something and I followed it, is the decision still mine?

Answer: Usually, yes — at least partly. The recommendation may come from the system, but your decision to rely on it is itself a decision.

The recommendation may come from the machine. The decision to trust the recommendation still belongs somewhere.

Advice does not automatically remove agency

We have always taken advice from other people.

A doctor recommends.

A lawyer advises.

A friend warns.

An analyst forecasts.

Advice can influence us strongly without becoming our action.

AI changes the scale and speed of advice, but not every principle of responsibility.

Ask what role AI actually played

Did AI give you information?

Did it recommend one option?

Did it rank several choices?

Did it make the choice automatically?

Did you have realistic freedom to reject it?

Responsibility changes depending on the answer.

“The AI told me” is not the same as “I had no choice”

If you had authority to decide and simply accepted the recommendation, you still participated in the outcome.

You may have relied reasonably on the system.

You may even have good reasons for relying on it.

But reasonable reliance is not the same as absence of responsibility.

The higher the stakes, the stronger the duty to review

Using AI to choose a restaurant is not the same as using AI to reject a job applicant.

Using AI to improve the wording of an email is not the same as using it to make a medical, legal, financial or disciplinary judgment.

The seriousness of the consequence should affect how much independent checking you owe.

Ask what the system cannot know

This question is essential because AI can appear complete while operating on an incomplete representation.

Confidence is not authority

An AI system may sound certain.

That does not give it moral authority.

Fluent language can make a recommendation feel more justified than it is.

Your job is to separate presentation from reason.

The author’s responsibility test

Four levels of reliance

• Informational: AI gives facts or options.
• Advisory: AI recommends a course of action.
• Default: AI’s recommendation is normally accepted unless challenged.
• Automatic: AI acts without case-by-case human approval.

Your responsibility changes as authority changes, but it rarely disappears completely if you chose the system or chose to rely on it.

What if the AI was wrong?

Then ask whether your reliance was reasonable.

Did you know the system’s limits?

Did you verify high-stakes information?

Did you ignore obvious warning signs?

Did you use AI in a domain where professional expertise was required?

What if the AI was better than me?

That is possible.

A tool can outperform you and still require responsible use.

Better prediction does not automatically answer every moral question about how the prediction should be used.

Do not outsource the final human question

Before acting, ask:

• Who could be harmed if this recommendation is wrong?
• What information might be missing?
• Would I be willing to explain this decision to the person affected?
• What reason do I have beyond “AI said so”?

The decision may be shared, but it is not ownerless

AI can influence a choice.

A company can design the system.

A manager can set the policy.

You can make the final click.

Responsibility may therefore be distributed among several people.

Shared responsibility is not no responsibility.

The safest question is not “Can AI decide for me?” It is “What responsibility do I still carry after I choose to let AI influence the decision?”

Research Context & References

1. Aristotle. Nicomachean Ethics, Book III.
2. Jonas, Hans. The Imperative of Responsibility. 1979.
3. Shahzad, Syed Raheel. Official Research programme, 2026.

Syed Raheel Shahzad

سيد راحيل شهزاد

Urdu: سید راحیل شہزاد · Hindi: सैयद राहील शहज़ाद

Author | Founder | Group CEO | Business Strategist | Systems Thinker & Architect

SyedRaheelShahzad.com · Research · Publications · Google Scholar · PhilPeople

Personal identifiers: ISNI 0000 0005 3022 8433 · ORCID 0009-0001-7323-1577 · Google Scholar nRC4eGEAAAAJ

Publisher / Imprint: The Syed Group
TheSyedGroup.com
Institutional ISNI: 0000 0005 3027 5408
Ringgold ID: 850493