Dave Turnbull, the Deputy Head of Educator AI Training at Educate Ventures Research, shares insights on the importance of risk assessments for AI learning tools in education. He discusses who these assessments are for and their accessibility for educators. The conversation highlights ethical risks, data privacy, and the need for clear communication about AI tools. They explore the challenges educators face when integrating AI while ensuring student safety, emphasizing the shared responsibilities of developers and educators in fostering effective use of AI technologies.
36:54
forum Ask episode
web_stories AI Snips
view_agenda Chapters
auto_awesome Transcript
info_circle Episode notes
volunteer_activism ADVICE
Use Concise Risk Tables
Read concise risk tables to quickly spot data and ethical concerns without technical expertise.
Use the mitigation tips and EVR staff notes to translate risk findings into classroom practice.
insights INSIGHT
Differentiate Data And Ethical Risks
Separate data risks (technical) from ethical risks (learner-focused) to clarify responsibilities.
Education-focused risk frames make assessments more relevant to teachers than generic AI risk tools.
question_answer ANECDOTE
Developer's Missed Risks Shaped The Work
Ibrahim recounts building an AI tutor and missing many practical risks during development.
That experience motivated EVR's risk assessments to surface overlooked privacy and bias issues.
Get the Snipd Podcast app to discover more snips from this episode
In today’s episode, we have the first part of a two-part miniseries on risk management, risk mitigation and risk assessment in AI learning tools. Professor Rose Luckin is away in Australia, speaking internationally, so Rowland Wells takes the reins to chat with Educate Ventures Research team members about their experience managing risk as teachers and developers. What does a risk assessment look like and whose responsibility is it to take onboard its insights? Rose joins our discussion group towards the end of the episode, and in the second instalment of the conversation, Rowland sits down with Dr Rajeshwari Iyer of sAInaptic to hear her perspective on risk and testing features of a tool as a developer and CEO herself.
Dave Turnbull, Deputy Head of Educator AI Training, EVR
Ibrahim Bashir, Technical Projects Manager, EVR
Rose Luckin, CEO & Founder, EVR
Talking points and questions include:
Who are these for? what’s the profile of the person we want to engage with these risk assessments? They’re concise, easy-to-read, no technical jargon. But it’s still an analysis, for people with a research/evidence mindset. Many people ignore it: we know that even learning tool developers who put research on their tools ON THEIR WEBSITES do not actually have it read by the public. So how do we get this in front of people? Do we lead the conversation with budget concerns? Safeguarding concerns? Value for money?
What’s the end goal of this? Are you trying to raise the sophistication of conservation around evidence and risk? Many developers who you critique might just think you’re trying to make a name pulling apart their tools. Surely the market will sort itself out?
What’s the process involved in making judgements about a risk assessment? If we’re trying to demonstrate to the buyers of these tools, the digital leads in schools and colleges, what to look for, what’s the first step? Can this be done quickly? Many who might benefit from AI tools might not have the time to exhaustively hunt out all the little details of a learning tool and interpret them themselves?
Schools aren’t testbeds for intellectual property or tech interventions. Why is it practitioners’ responsibilities to make these kind of evaluations, even with the aid of these kind of assessments? Why is the tech and AI sector not capable of regulating their own practices?
You’ve all worked with schools and learning and training institutions using AI tools. Although this episode is about using the tools wisely, effectively and safely, please tell us how you’ve seen teaching and learning enhanced with the safe and impactful use of AI
The Edtech Podcast