Respect for Persons and Human Dignity
A central tenet of ethics is respect for the intrinsic value of persons, or human dignity The human right to be valued and treated with respect because of one's personhood. in all interactions. Respect comprises:
Including:
- Regard for their welfare
- Consideration of the views, norms, and cultural context of individuals involved in research and the communities from which they are recruited (and not recruited)
- Respect for privacy and confidentiality of participants and their communities
- Consideration of the capacity of humans to make their own decisions, and empowering those who may have diminished capacity to make such decisions, with particular care towards these vulnerable populations
Autonomy is normally taken to comprise:
- informed consent - that participants may choose to freely participate or not, based on an understanding of that participation
- agency - that people have the opportunity to participate in research, that their views are heard, respected, and fairly represented, both individually and with respect to cultural norms and values.
Autonomy is about having freedom from external control, being able to according to one's own values; it is fundamental to informed consent. Agency is about one's capacity to be heard and play an active role, recognising participants as active contributors to research.
In the context of emerging technologies, these ethical concerns involve...
here to review
check this, these are now drawn from the same table as the concepts in the introduction and so the thread should be clear.
How is AI research different?
ChallengeInstances
Respect for human-human relationships, at individual and collective levels
There is potential for breakdown in trust to arise from a loss of exposure to diverse human-human interactions
“While the capacity of AI systems to curate individual experiences and to personalise digital services holds the promise of vastly improving consumer life and service delivery, this benefit also comes with potential risks. Excessive automation, for example, might reduce the need for human-to-human interaction, while algorithmically enabled hyper...
ChallengeInstances
Privacy, confidentiality, data integrity and cultural norms
It is not always clear how we can protect our identities to assure privacy and identity verification
"**Issue: **How can we increase agency by providing individuals access to services allowing them to create a trusted identity to control the safe, specific, and finite exchange of their data?## BackgroundPervasive behavior-tracking adversely affects human agency by recognizing our identity in every action we take on and offline. This is why iden...
ChallengeInstances
Effective use of AI inherently involves multiple stakeholder and disciplinary expertise, including both AI literacy and contextual awareness across cultures
Defining desired outcomes - including general wellbeing - requires stakeholder involvement
“AI technologies are grounded in models, and these models are inevitably incomplete in some way. It is up to humans to name educational goals and measure the degree to which models fit and are useful—or don’t fit and might be harmful. Such an assessment of how well certain tools serve educational priorities may seem obvious, but the romance of t...
ChallengeInstances
Respect for diversity, non-discrimination, and fairness
How do we ensure protection of data during humanitarian emergencies?
"\*_"Issue: \*_A/IS are contributing to humanitarian action to save lives, alleviate suffering, and maintain human
dignity
The human right to be valued and treated with respect because of one's personhood.
both during and in the aftermath of man-made crises and natural disasters, as well as to prevent and strengthen preparedness for the occurrence of such situations. However, there are ethical concerns with both the collection...
ChallengeInstances
Human control of technology and accountability for technology use
How do we provide consistent oversight of AI to ensure they are accountable to end-users for any conclusions?
**Issue:** Oversight for algorithms## BackgroundThe algorithms behind A/IS are not subject to consistent oversight. This lack of assessment causes concern because end users have no account of how a certain algorithm or system came to its conclusions. These recommendations are similar to those made in the “General Principles” and “Embedding Value...
ChallengeInstances
Privacy, confidentiality, data integrity and cultural norms
It is not always clear how our data is being used, or how we could find out
It is not always clear what data about is being used, how it is being used, by whom, and to what end. "Issue: What would it mean for a person to have an algorithmic agent helping them actively represent and curate their terms and conditions at all times?While it’s essential to create your own terms and conditions to broadcast your preferences, i...
ChallengeInstances
Respect for diversity, non-discrimination, and fairness
Implementation fairness and bias in system use, where undue dis/trust is placed in automated systems including in ways that are inequitable, or/and that deskill professionals
"Implementation fairness When your project team is approaching the beta stage, you should begin to build out your plan for implementation training and support. This plan should include adequate preparation for the responsible and unbiased deployment of the AI system by its on-the-ground users. Automated decision-support systems present novel ris...
ChallengeInstances
Commercial interests - while significant AI work is derived from publicly funded research, commercial actors have significant interests and this may impact relationships between funders-technology providers-researchers-public with blured lines between these groups, and complex power dynamics
There may be lack of accountability for centering stakeholder experience and agency in the design of AI driven by commercial interests
AI systems may impact on relationships between decision makers such that judgements previously made at a professional level must involve technology providers. It may not be clear in advance how the user experience of AI tools has been considered in evaluation, nor what input those using the tool can have into ongoing development. This may occur ...
ChallengeInstances
Interactions and distance between researchers and participants are central to research, both when proximal, and when mediated by technologies.
Technologies mediate the interactions and distance between researchers and participants
Tools may mediate the interaction between researchers and participants, or research may be conducted in a way that is distal to human interactions with the inputs or outputs of the research (e.g., distal to the original capture of the data in the context of gathering training data; or to the use of a tool in the context of training an algorithm ...
ChallengeInstances
Privacy, confidentiality, data integrity and cultural norms
Non-research platforms such as smart toys may collect data about children with little regulation
**Issue: **Intelligent toys## BackgroundChildren will not only be exposed to A/IS at school but also at home, while they play and while they sleep. Toys are already being sold that offer interactive, intelligent opportunities for play. Many of them collect video and audio data which is stored on company servers and either is or could be mined fo...
ChallengeInstances
Respect for human-human relationships, at individual and collective levels
Management may experience reduced autonomy, through automisation that limits creative, affective, empathetic concerns
"Will extensive use ofA/IS in society make our organizations more brittle by reducing human autonomy within organizations, and by replacing creative, affective, empathetic componentsof management chains?## BackgroundIf human workers are replaced by A/IS, the possibility of corporations, governments, employees, and customers discovering new equil...
ChallengeInstances
Privacy, confidentiality, data integrity and cultural norms
Generic T&C statements provide limited control to individuals over their own data
"**Issue**: What would it mean for a person to have individually controlled terms and conditions for their personal data?## BackgroundPart of providing individually controlled terms and conditions for personal data is to help each person consider what their preferences are regarding their data versus dictating how they need to share it. While qu...
ChallengeInstances
Effective use of AI inherently involves multiple stakeholder and disciplinary expertise, including both AI literacy and contextual awareness across cultures
Norms, aims, and practices are diverse and may vary by location, change over time, and come into conflict
A/IS culture and context## Norms vary: BackgroundA responsible approach to embedding values into A/IS requires that algorithms and systems are created in a way that is sensitive to the variation of ethical practices and beliefs across cultures. The designers of A/IS need to be mindful of cross-cultural ethical variations while also respecting wi...
ChallengeInstances
Respect for human-human relationships, at individual and collective levels
How might AI deployed in care settings to foster intimate relationships impact on relationships among humans?
Are moral and ethical boundaries crossed when the design of affective systems allows them to develop intimate relationships with their users?## BackgroundThere are many robots in development or production designed to focus on intimate care of children, adults, and the elderly2. While robots capable of participating fully in intimate relationship...
ChallengeInstances
Respect for human-human relationships, at individual and collective levels
Access to increasing amounts of data about each other may impact our interactions in ways that change human autonomy
"Does the increased access to personal information about other members of our society, facilitated by A/IS, alter the human affective experience?Does this access potentially lead to a change in human autonomy?## BackgroundTheoretical biology tells us that we should expect increased communication—which A/IS facilitate— to increase group-level inv...
ChallengeInstances
Commercial interests - while significant AI work is derived from publicly funded research, commercial actors have significant interests and this may impact relationships between funders-technology providers-researchers-public with blured lines between these groups, and complex power dynamics
Implementation of AI may be driven by commercial, not values-based, aims
### **Values-based ethical culture and practices for industry**Corporations are built to create profit while competing for market share. This can lead corporations to focus on growth at the expense of avoiding negative ethical consequences. Given the deep ethical implications of widespread deployment of A/IS, in addition to laws and regulations,...
ChallengeInstances
Effective use of AI inherently involves multiple stakeholder and disciplinary expertise, including both AI literacy and contextual awareness across cultures
Agency may be stymied in the context of black box models
"Use of black-box componentsSoftware developers regularly use “black box” components in their software, the functioning of which they often do not fully understand. “Deep” machine learning processes, which are driving many advancements in autonomous and intelligent systems, are a growing source of black box software. At least for the foreseeable...
ChallengeInstances
Research produces both direct impact, on participants and then society, while also giving rise to indirect, long-range, and dual-use impacts
Respect for human-human relationships, at individual and collective levels
AI has the potential to impact on how we connect and talk together, and the structure of societies. AI should avoid harms and promote pro-social engagement, recognising that both the benefits and harms of research (and broader engagement with technology) may accrue over groups.
ChallengeInstances
Respect for human-human relationships, at individual and collective levels
Longstanding concern regarding 'deception' should be applied to the context of affective AI systems
"When, if ever, andunder which circumstances,is deception performed by affective systems acceptable?## BackgroundDeception is commonplace in everyday human-human interaction. According to Kantian ethics, it is never ethically appropriate to lie, while utilitarian frameworks indicate that it can be acceptable when deception increases overall happ...
And for education
here to review
check this, these are now drawn from the same table as the concepts in the introduction and so the thread should be clear
How is education research different?
ChallengeInstances
Effective use of AI inherently involves multiple stakeholder and disciplinary expertise, including both AI literacy and contextual awareness across cultures
Defining desired outcomes - including general wellbeing - requires stakeholder involvement
“AI technologies are grounded in models, and these models are inevitably incomplete in some way. It is up to humans to name educational goals and measure the degree to which models fit and are useful—or don’t fit and might be harmful. Such an assessment of how well certain tools serve educational priorities may seem obvious, but the romance of t...
ChallengeInstances
Interactions and distance between researchers and participants are central to research, both when proximal, and when mediated by technologies.
Research involving learning may seek to draw on secondary analysis of learner data
Including the principles (which are drawn from the Canadian model)## TCPS2 | “Reasons to Conduct Secondary Analyses of Data”“Reasons to conduct secondary analyses of data include: avoidance of duplication in primary collection and the associated reduction of burdens on participants; corroboration or criticism of the conclusions of the original p...
ChallengeInstances
Commercial interests - while significant AI work is derived from publicly funded research, commercial actors have significant interests and this may impact relationships between funders-technology providers-researchers-public with blured lines between these groups, and complex power dynamics
There may be lack of accountability for centering stakeholder experience and agency in the design of AI driven by commercial interests
AI systems may impact on relationships between decision makers such that judgements previously made at a professional level must involve technology providers. It may not be clear in advance how the user experience of AI tools has been considered in evaluation, nor what input those using the tool can have into ongoing development. This may occur ...
ChallengeInstances
Privacy, confidentiality, data integrity and cultural norms
Non-research platforms such as smart toys may collect data about children with little regulation
**Issue: **Intelligent toys## BackgroundChildren will not only be exposed to A/IS at school but also at home, while they play and while they sleep. Toys are already being sold that offer interactive, intelligent opportunities for play. Many of them collect video and audio data which is stored on company servers and either is or could be mined fo...
ChallengeInstances
Interactions and distance between researchers and participants are central to research, both when proximal, and when mediated by technologies.
Research involving teacher-student relationships involves power dynamics
Research in learning settings may involve research in the tradition of scholarship of teaching and learning (SoTL) research, or other 'close to practice' approaches (e.g. action research, design research, etc.). In this kind of research there are power dynamics between researchers who may be both researcher and teacher, placing them in a positi...
ChallengeInstances
Research produces both direct impact, on participants and then society, while also giving rise to indirect, long-range, and dual-use impacts
Reification, ossification, and standardisation reduce autonomy
Use of AI in education risks a set of long-range and indirect impacts that may be difficult to assess. These include the risks of AI1. Reifying assumptions, acting to label learners and teachers and thus shape their lifelong trajectories, rather than providing useful information to support agency in shaping those trajectories. Data in learning e...
ChallengeInstances
Privacy, confidentiality, data integrity and cultural norms
Generic T&C statements provide limited control to individuals over their own data
"**Issue**: What would it mean for a person to have individually controlled terms and conditions for their personal data?## BackgroundPart of providing individually controlled terms and conditions for personal data is to help each person consider what their preferences are regarding their data versus dictating how they need to share it. While qu...
ChallengeInstances
Effective use of AI inherently involves multiple stakeholder and disciplinary expertise, including both AI literacy and contextual awareness across cultures
Agency may be stymied in the context of black box models
"Use of black-box componentsSoftware developers regularly use “black box” components in their software, the functioning of which they often do not fully understand. “Deep” machine learning processes, which are driving many advancements in autonomous and intelligent systems, are a growing source of black box software. At least for the foreseeable...
ChallengeInstances
Interactions and distance between researchers and participants are central to research, both when proximal, and when mediated by technologies.
Data collection or/and analysis often involves methods that challenge individually based participation models
In the content of research there are often challenges in disaggregating data. For example, consider classroom based research in which (1) there are challenges in excluding students from capture in video data, and where all students are using the same learning platform (which also provides research data); and (2) where group-work or whole-class f...
Technical procedural ethics addressing respect for persons
to add here
tbc. May link to the principles drawn on from the principlemapping table
This section draws on principles discussed in
Sources
EC guidelines on AI in education for educators
“These Guidelines are to be used in schools across Europe...
AI education-research ethics-guideline not-research-ethics
Sources
Khan-review systematic literature analysis
A review piece
Sources
Schiff-review AI Ethics Global Document Collection
A review piece
Sources
Jobin-review global landscape of AI ethics guidelines
A review piece
Sources
National Statement on Ethical Conduct in Human Research
"The **National Statement** is intended for use by:- any ...
ethics-guideline research-ethics
Sources
Turing responsible design and implementation of AI systems in the public sector
"The Public Policy Programme at The Alan Turing Institute...
AI ethics-guideline not-research-ethics
Sources
High-level Expert Group on AI
“These guidelines are addressed to all AI stakeholders de...
AI ethics-guideline not-research-ethics
Sources
IEEE
"Autonomous and intelligent technical systems are specifi...
AI ethics-guideline research-ethics
Sources
Fjeld-review mapping consensus in AI ethics
This section draws on
Sources
EC guidelines on AI in education for educators
“These Guidelines are to be used in schools across Europe...
AI education-research ethics-guideline not-research-ethics
Sources
Ethics in The Scholarship of Teaching and Learning
“This new Taylor Institute Guide takes the researcher thr...
education-research ethics-guideline research-ethics
Sources
National Statement on Ethical Conduct in Human Research
"The **National Statement** is intended for use by:- any ...
ethics-guideline research-ethics
Sources
Turing responsible design and implementation of AI systems in the public sector
"The Public Policy Programme at The Alan Turing Institute...
AI ethics-guideline not-research-ethics
Sources
IEEE
"Autonomous and intelligent technical systems are specifi...
Particular Strategies
These are strategies that relate to respect for persons.
Strategies
Overview of considerations in transparency
“Requirements of Trustworthy AI The principles outlined i...
Strategies
Questions to consider in key stages of AI and machine learning based research, regarding legitmacy and power in consent
“Conducting research, especially with complex and often o...
project-design reflection-questions
Strategies
Principle of Accountability, key considerations
"When considering the role of accountability in the AI pr...
Strategies
Educate the learners in formal education, the public, professionals, and policy makers for designing, and working alongside, AI to advance the SDGs.
## "RecommendationsEducation with respect to A/IS must be...
Strategies
Guiding questions for educators regarding Diversity, non-Discrimination, and Fairness of AI in Education
- “Is the system accessible by everyone in the same way w...
Strategies
Questions to consider in key stages of AI and machine learning based research, regarding sociotechnical context
“The first step is to create an understanding of the AI, ...
project-inception reflection-questions
Strategies
How do we navigate individual choice of technology engagement and school or system-wide procurement?
"Local and national educational authorities must work to ...
Strategies
Considerations in assessing trustworthy AI - Access to data
“Compliance with this assessment list is not evidence of ...
Strategies
Questions to consider in key stages of AI and machine learning based research, regarding transparency and explainability
The scientific method requires a high degree of transpare...
governance-question project-dissemination reflection-questions
Strategies
Systematic analysis of consequences, transparency, user agency, and safeguards are central to implementation of AI nudges
## "Recommendations1\.Systematic analyses are needed that...
Strategies
Principle of Discriminatory non-harm for fairness, key considerations for implementation fairness
"Implementation fairness When your project team is approa...
Strategies
Questions to consider in key stages of AI and machine learning based research, regarding data collection
“Data collection for the model training and the research ...
project-design reflection-questions
Strategies
Considerations in assessing trustworthy AI - Communication
“Compliance with this assessment list is not evidence of ...
Strategies
Key discussion points for planning SoTL research regarding power relationships
## Conflicts of Interestand Power Relationships“**Key Pri...
education-research reflection-questions
Strategies
AI assistants should be created to help users understand and manage how their data is being used
"Algorithmic agents should be developed for individuals t...
Strategies
Consider tensions and absolute rights
“Tensions may arise between the above principles, for whi...
Strategies
Key discussion points for planning SoTL research regarding secondary analysis of data
Including the principles (which are drawn from the Canadi...
education-research reflection-questions
Strategies
Questions to consider regarding re-identification and issues of justice in IP
“How are findings presented? - What immediate or future r...
Strategies
Considerations in assessing trustworthy AI - Minimising and reporting negative impacts
“Compliance with this assessment list is not evidence of ...
Strategies
Consider developing strategies for specific context of humanitarian action
See also \[Engine Room report In the humanitarian sector’...
Strategies
Questions to consider regarding access to a context (and data) and the perceptions and autonomy of those related to that context regarding access to it
- "How is the context (venue/participants/data) being acc...
Strategies
Considerations in assessing trustworthy AI - Human oversight
“Compliance with this assessment list is not evidence of ...
Strategies
Principle of Discriminatory non-harm for fairness, key considerations for outcome fairness
"Once you and your project team have thoroughly considere...
Strategies
Guiding questions for educators regarding Privacy and Data Governance of AI in Education
- "Are there mechanisms to ensure that sensitive data is ...
Strategies
Questions to consider regarding protection of vulnerable populations
“What particular issues might arise around the issue of m...
Strategies
Considerations in assessing trustworthy AI - Stakeholder participation
“Compliance with this assessment list is not evidence of ...
Strategies
Guiding questions for educators regarding Accountability of AI in Education
- "Who is responsible for the ongoing monitoring of resul...
Strategies
Considerations in assessing trustworthy AI - Unfair bias avoidance
“Compliance with this assessment list is not evidence of ...
Strategies
Foundational issue: Promotion of Transparency
“AI models in edtech will be approximations of reality an...
Strategies
Always Center Educators in Instructional Loops
“To succeed with AI as an enhancement to learning and tea...
Strategies
Questions to consider in assessing risk of harms
“What are the potential harms or risks associated with th...
Strategies
Wellbeing measures and promotion should be part of AI evaluation
## "RecommendationsA/IS creators should work to better un...
Strategies
Questions to consider regarding autonomy and informed consent
“How are we recognizing the autonomy of others and acknow...
Strategies
Considerations in assessing trustworthy AI - Documenting trade-offs
“Compliance with this assessment list is not evidence of ...
Strategies
Community norms should be identified, and expertise from intercultural information ethics practitioners embedded in ethics committees
## RecommendationEstablish a leading role for [intercultu...
Strategies
Considerations in assessing trustworthy AI - Traceability
“Compliance with this assessment list is not evidence of ...
Strategies
Guiding questions for educators regarding Human Agency and Oversight of AI in Education
- Is the teacher role clearly defined so as to ensure tha...
Strategies
Questions to consider regarding the norms and values of those involved in the study and alignment of these to the research
- “Who is involved in the study? - What are the ethical e...
Strategies
Considerations in assessing trustworthy AI - Society and democracy
“Compliance with this assessment list is not evidence of ...
Strategies
Training in skills for adaptability to rapid technological changes informed by improved data regarding labour pattern shifts
## "RecommendationsTo thrive in the A/IS age, workers mus...
Strategies
Questions to consider regarding data management and participant re-identification
“How are data being managed, stored, and represented? - W...
Strategies
Principle of transparency, key considerations
"It is important to remember that _transparency as a prin...
Strategies
Principle of Discriminatory non-harm for fairness, key considerations for design fairness
“Because human beings have a hand in all stages of the co...
Strategies
Guiding questions for educators regarding Transparency of AI in Education
- "Are teachers and school leaders aware of the AI method...
Strategies
Questions to consider in SoTL research inception
“Purpose, Participation, and Consent - What is the questi...
education-research reflection-questions
Strategies
Considerations in assessing trustworthy AI - Quality and integrity of data
“Compliance with this assessment list is not evidence of ...
Strategies
Educational data should be classified as sensitive and held in 'escrow' not available for commercial purposes
"Educational data offer a unique opportunity to model ind...
Strategies
Principle of Discriminatory non-harm for fairness, key considerations for data fairness
"To ensure the uptake of best practices for responsible d...
Strategies
Questions to consider in key stages of AI and machine learning based research, regarding the Research Analytical Process
"The research analytical process includes selecting the t...
project-analysis reflection-questions
Strategies
Questions to consider regarding how a context is defined and conceptualised with respect to relevant stakeholder agency
“How is the context defined and conceptualized? - Does th...
Strategies
Questions to consider regarding the nature of the data (as (dis)aggregated, private/public, reidentifiable, etc.)
“What is the primary object of study? - What are the ethi...
Strategies
Considerations in assessing trustworthy AI - Accessibility and universal design
“Compliance with this assessment list is not evidence of ...
Strategies
Consider where AI-mediation is, and is not, appropriate with respect to human relationships and autonomy
## "Recommendations1\.It is important that human workers’...
Strategies
Considerations in assessing trustworthy AI - human agency
“Compliance with this assessment list is not evidence of ...
Strategies
Tools for individuals to create custom machine-readable dynamic terms and conditions that respect their preferences for data collection and use
"Individuals should be provided tools that produce machin...
Strategies
Considerations in assessing trustworthy AI - Respect for privacy and data protection
“Compliance with this assessment list is not evidence of ...
Strategies
Considerations in assessing trustworthy AI - Auditability
“Compliance with this assessment list is not evidence of ...
Strategies
Questions to consider regarding re-identification and disciplinary methodological norms
Particular Cases
Cases
Using chatbots to guide learners and parents through administrative tasks
“A school uses a chatbot virtual assistant on its website...
Resource-management-and-administration
Cases
Providing individualised interventions for special needs
“A school is considering how AI systems can help reduce b...
Dropout-risk-and-grade-prediction
Cases
Managing student enrolment and resource planning
A school uses the data collected when students enrol to p...
Resource-management-and-administration
Cases
Teacher evaluation fails to acknowledge individual contexts and teacher autonomy
Cases
Using adaptive learning technologies to adapt to each learner’s ability
“A primary school is using an Intelligent Tutoring System...
Adaptive-and-personalised-learning
Cases
Automated Essay Scoring is proposed in order to reduce the marking burden, improve feedback turnaround, and standardise outcomes
“A school is looking at how AI systems can support the as...
Cases