vulnerable people and their data being shared across social media & opening them up to identification- question this bigger issue effects children and across society- what are we doing about this?

Hearing from a software developer in the audience raises several points 1) schools know as much about data protection as per software developers- very little- how to anonymise, pseudonoymise, encryption 2) private companies protect data for reputational risk rather than GDPR 3) flagging of

Suggestions of amendments- could we look at the existing Children Code to expand remit e.g enabling EdTech delivered in schools to be included? What can we learn internationally of lessons learned. Should it be positioned as how we enhance missions? But also need to bring back to data protection

A peer reminds us of the importance for civil society organisations to work together to develop joint briefs of clear actions as the Data Use & Access Bill goes through the parliamentary stages. The House of Lords provides an opportunity for clear case to be made

Dr Frabboni recommends increased literacy around copyright is needed. We also need to discuss what do children as authors want? It might not be renumeration. @theabb.bsky.social reflects this feels new to think about childrens content as open to being commercialised

Considering the art & creativity children produce- do parents consider themselves as stewards of children's content? Should children's creativity be used freely to develop generative AI apps (as part of the social agreement of being provided an education)?

@theabb.bsky.social raises a number of reports that show people don't want personal sensitive data to be pooled into lakes to be used for any purposes. Now to discuss work from Dr Maria Frabboni looking at copyright & IP of work of vulnerable authors & tonight the work produced by children

None of the parents or teachers we spoke to mentioned any reviews conducted by the schools about the efficacy of these apps. More findings to come, but Ade's concludes that with the data use and access bill potential weakening rights on purpose limitation, legitimate interest, is concerning

Ade bring forward the lead: Being an effective guardian of a child's data is not an easy thing for parents and guardians living in England and Wales. Most parents/carers felt uninformed on how childrens data is used in EdTech apps.

Research with ten community researchers- 90 school websites reviewed examining mentions of EdTech, privacy policies & equality statements, 28 interviews, 67 responses to questionnaires + community events.

Hearing from my colleague Ade Adewunmi at Data, Tech & Black Communities sharing findings from the community research project we've worked on looking at EdTech use in Birmingham primary & secondary schools & specifically with Black communities. Report will be out next month.

Clear guideline required on how to collect consent, to explain risks and to explain how to limit exposure of images e.g making pages private- which has some flaws but is a step; never publishing a child's name- but face linked to a school makes kids identifier

Dr Claire Bessant's research highlight a number of serious questions to be investigated, in some places, declining parental consent for children's images to be taken meant their children couldn't participate in activities! 19 authorities collected a blank option for general consent for everything!

Good grief, just hearing that images of children shared my parents and schools are being used in large training datasets (where children can be identifiable), can be used for deep fakes and finding images in a range of different databases

Oh don’t get me started on social care (elderly father & disabled brother). There are so many basic infrastructure work required there that anytime I see the word AI I feel like a cartoon character with flames shooting out of my ears.

& I want more details- how many times did the AI spot things a human didn’t? And how many times vice versa? How accurate are the summary notes & how many times does it give harmful advice? Who exactly is doing the evaluations?…. Not just little snippets saying how great the tool is