Short answer

When designing AI systems, dedicate specific design efforts and evaluation criteria to ethical principles, in addition to user interaction and usability.

Field
User-Centred Design
Source
Academic Publication (2022)
Method
Co-evaluation workshop
Sample
4 participants
Evidence
Moderate effect

Designing for effective human-AI interaction does not automatically ensure that an AI system adheres to ethical principles. This user-centred design research insight is drawn from a 2022 study published in Academic Publication. Using Co-evaluation workshop with 4 participants, researchers explored how this design variable affects real-world outcomes. The key design takeaway: When designing AI systems, dedicate specific design efforts and evaluation criteria to ethical principles, in addition to user interaction and usability.

Study
User-Centred DesignHigh ImpactModerate effect

Human-AI Interaction Guidelines Don't Guarantee Ethical AI Design

Designing for effective human-AI interaction does not automatically ensure that an AI system adheres to ethical principles.

Academic Publication · 2022

01

Key Findings

  • 01Compliance with human-AI interaction guidelines did not necessarily correlate with compliance with ethical AI guidelines.
  • 02Compliance with ethical AI guidelines did not necessarily correlate with compliance with human-AI interaction guidelines.
02

Application

Design takeaway

When designing AI systems, dedicate specific design efforts and evaluation criteria to ethical principles, in addition to user interaction and usability.

How to apply

When developing AI products, use checklists or frameworks that address both user experience (e.g., Microsoft's HAI guidelines) and ethical considerations (e.g., EU's Ethics Guidelines for Trustworthy AI) independently.

Project actions

  • 01When evaluating AI products for your design project, consider using separate rubrics for user interaction and ethical compliance.
  • 02Look for examples where good user experience might conflict with ethical outcomes, or vice versa.
03

Method & Evidence

AimTo investigate the extent to which existing AI-enabled applications comply with both human-AI interaction guidelines and ethical AI guidelines.
MethodCo-evaluation workshop
ProcedureFour participants with AI design experience evaluated two existing AI-enabled applications (Strava and CoronaMelder) for their adherence to both human-AI interaction guidelines and ethical AI guidelines.
Sample4 participants
ContextDesign of AI-enabled applications

Variables

IVAdherence to human-AI interaction guidelines, Adherence to ethical AI guidelines
DVCompliance of AI-enabled applications with guidelines
CVType of AI-enabled application, Participants' prior knowledge of AI design
04

Strengths & Limitations

Strengths

  • +Focuses on a critical and under-discussed aspect of AI design.
  • +Uses a practical case study approach to investigate real-world applications.

Limitations

The subjective nature of ethical judgments and the specific context of the chosen applications might influence the results.

Reliability & validity

The reliability of the findings could be improved with a larger sample size of participants and applications. Validity is supported by the use of established guideline sets for evaluation.

Think critically

How might the perceived 'trustworthiness' of an AI system be affected by a conflict between its interaction design and its ethical framework?

05

Design Principles

"Ethical AI design is a distinct requirement from effective human-AI interaction design."

This finding highlights a critical gap in the design process for AI-enabled products. Designers must consider ethical implications as a distinct and parallel concern to usability and interaction flow, rather than assuming they are inherently addressed by good HCI practices.

06

What This Means for Your Design

Just because an app is easy to use with AI doesn't mean it's fair or trustworthy. You need to check both things separately.

How to use in your project

  • 1.Reference this study when discussing the importance of a dual approach to evaluating AI systems in your design project's analysis or evaluation sections.
07

Add to My Project

08

Quick Cite

Paragraph starter

The research by Li and Lü (2022) indicates that adherence to human-AI interaction guidelines does not guarantee compliance with ethical AI principles. This suggests that designers must adopt a dual approach, evaluating both usability and ethical implications independently to ensure the responsible development of AI-enabled solutions.

09

Source

Academic Publication

Human-AI interaction and ethics of AI: how well are we following the guidelines

journal · 2022

View source

Questions About This Research

What does the research say about human-ai interaction guidelines don't guarantee ethical ai design?
When designing AI systems, dedicate specific design efforts and evaluation criteria to ethical principles, in addition to user interaction and usability. Evidence: Academic Publication (2022).
Why does "Human-AI Interaction Guidelines Don't Guarantee Ethical AI Design" matter for design?
This finding highlights a critical gap in the design process for AI-enabled products. Designers must consider ethical implications as a distinct and parallel concern to usability and interaction flow, rather than assuming they are inherently addressed by good HCI practices.
How can designers apply this research?
When designing AI systems, dedicate specific design efforts and evaluation criteria to ethical principles, in addition to user interaction and usability.
What were the main findings?
Compliance with human-AI interaction guidelines did not necessarily correlate with compliance with ethical AI guidelines.. Compliance with ethical AI guidelines did not necessarily correlate with compliance with human-AI interaction guidelines.
What research method was used?
Co-evaluation workshop with 4 participants.
How strong is the evidence?
Evidence strength is rated Moderate effect, based on a 2022 journal from Academic Publication.
What should I do differently in my next project?
When developing AI products, use checklists or frameworks that address both user experience (e.g., Microsoft's HAI guidelines) and ethical considerations (e.g., EU's Ethics Guidelines for Trustworthy AI) independently.
What are the limitations?
The study involved a small number of participants and focused on only two specific applications, limiting the generalizability of the findings.