Part 1 · Know Your AI Readiness
Ethics · Spotting Ethical Issues in AI
The Mindset lessons taught you to reach for AI on new problems. So you hand it a real one: shortlist ten CVs from a stack of two hundred. It returns a clean list in seconds, and every candidate on it fits the same profile. Something is off. What?
Where ethical issues hide
That clean list is the problem. The AI learned from who your organisation hired before. If past hiring leaned toward one profile, the tool now reads that profile as the definition of a strong candidate, and quietly screens out the rest. There is no malice in it. It is faithfully repeating a skewed past. That is bias: a pattern baked into the training data, inherited without anyone choosing it.
Most ethical issues in everyday AI look like this: a quiet skew, or a quiet exposure, that lands on a real person. Three lenses help you name one.
- Bias: whose data trained it? Swap a detail and see if the answer moves.
- Privacy: does this expose data that belongs to someone else?
- Accountability: the AI does not own the outcome. You do.
First, bias. Ask whose data trained the output, and whether that data could carry a skew. A quick test, the swap test: change one detail that should not matter, a name or a postcode, and run it again; if the answer shifts, you have found bias. Second, privacy. Ask whether the tool exposes data that belongs to someone else. Paste a customer's record into a public chatbot, and you may have handed it to a system that keeps it. Third, accountability. If the shortlist is wrong, the hiring manager answers for it. The tool answers for nothing.
Key Takeaway
The output that looks cleanest is often the one repeating a skew most confidently. That is the one to check, not the one to trust.
Check your understanding
Self-check only. Nothing is scored and nothing is recorded; open an answer to compare it with your own.
A team uses an AI tool to screen job applications and shortlist candidates. Which response best reflects what AI Ready looks like on ethics?
- Trust the shortlist, the AI is objective because it is a machine, so it removes human bias.
- Check whether the shortlist systematically favours or excludes particular groups, because the tool learned from historical decisions that may carry bias.
- Avoid AI for hiring entirely, since any use is unethical.
- Ethics is the legal team's job, so use the shortlist and let them flag problems later.
Show the answer
Check whether the shortlist systematically favours or excludes particular groups, because the tool learned from historical decisions that may carry bias.
Which of these is the clearest sign someone is below AI Ready on ethics?
- They check AI outputs they rely on for signs of unfairness before acting.
- They can point to who an AI decision affects and whether it is fair.
- They have heard the terms 'AI bias' and 'fairness' but have never applied them to anything they actually use.
- They raise fairness concerns with colleagues who can act on them.
Show the answer
They have heard the terms 'AI bias' and 'fairness' but have never applied them to anything they actually use.
Saved in this browser only. Carry progress between devices with the resume link on the course page.