During CIRCLS AI and Education Policy working group, we created an AI in Educational Technology Vendor Pledge to encourage more ethical technology development.
Why a Pledge?
There’s a continuum of recommendation to pledge to certification to mandate. Organizations on a national and global level have already put work into developing recommendations for ethical AI, both generally and specific to educational settings. Pledges take the next step of asking companies to commit to certain standards voluntarily and demonstrate commitment. Some pledges are legally enforceable once an organization signs on. Pledges do not necessarily require organizations to submit for evaluation as certifications do, and are not broadly mandatory like laws or policies.
When it comes to AI in educational settings, a pledge can help set the lowest bar that certifications and regulations can then build on. By developing this pledge, we are building on a growing body of educational technology pledges that ask edTech vendors to ensure transparency, privacy, data interoperability, and other important issues.
Who is our Audience?
We hope that this pledge can and will be signed by any educational technology companies that are developing tools that use artificial intelligence. We hope that this pledge will serve as a joint statement of values and set the lowest bar for acceptable use of AI in educational technology to improve education as a whole. Finally, we hope that we can provide sufficient incentive to companies to sign on to this pledge. If you are reading this document, we would appreciate your feedback on our pledge components, including anything that we may have missed or might be out of reach for EdTech companies at this time, and would appreciate feedback and ideas on how to make this pledge appealing and mutually beneficial to companies as well as educational institutions
Contributors:
Trisha Callella, Erin Walker, Leah Friedman, David Lockett, Lin Lin, Berj Akian
This pledge is a working document. We would like to formalize the sign on process and are open to feedback on next steps. Please email leahf@pitt.edu to contribute.
