WEEK 4, ARTICLE 2 Flashcards
(18 cards)
What’s the main issue the article identifies with AI ethics?
Many ethics guidelines exist, but REAL IMPLEMENTATION is weak — especially in AI systems.
What are 2 common problems with ethics frameworks?
Too abstract (e.g., fairness without how-to)
Leads to ethics washing (talk without action)
What is a key challenge with responsibility?
It’s often blamed on AI, not the humans who design or use it.
What is lacking in AI ethics tools?
Practical tools that help apply ethics in real product design.
How was the study done?
Survey of 54 practitioners
Interviews (startups, tech, public sector)
What 3 questions did the study ask?
How is ethics understood in practice?
What motivates ethics in design?
What support is needed for implementation?
What is a narrow view of ethical design?
Most focus on privacy/security, ignoring values like autonomy or solidarity.
Why is ethics often avoided in design?
Seen as too costly and with no clear ROI (return on investment).
What’s the problem with accountability?
Confusion over who is responsible
Practitioners feel unprotected and unsupported
No clear leadership commitment
Why do current ethics tools fall short?
They are too abstract and don’t give clear steps for real projects.
What is missing externally?
External checks like ethics audits — only 6% used them.
What logic do many companies use for AI ethics?
Risk-based logic (avoid legal trouble), not proactive ethical design.
What cultural shift is needed?
Move from abstract values to practical questions like:
“Is this the right solution?”
“Will this impact people positively?”
What is needed to make AI ethics real?
Combine legal compliance with ethical design
Build tools, accountability systems, and supportive environments
What kind of training should be added?
Mandatory AI ethics training (like in medicine), focusing on real-life decisions and empathy.
What’s one example framework mentioned?
Digital Catapult Ethics Framework (DCEF) – offers interim ethics tools, transparency sharing, and inclusive design.
What mindset should companies adopt for AI ethics?
Use iterative governance: keep improving with feedback
Involve marginalized groups
Focus on ethics, not just compliance
What’s the first step to better AI ethics?
Start now with small, practical steps — don’t wait for perfect solutions.
