American Federation of Teachers President Randi Weingarten Talks AI in Schools

SPEAKER_1 (0:02)

Bloomberg Audio Studios, Podcasts, Radio, News.

Carol Massar (0:07)

I want to bring in Randi Weingarten. She's president of the American Federation of Teachers. It's the second largest teachers union. 1.8 million members, including teachers, higher education faculty and staff, nurses and other health care professionals and more. She joins us from New York. Randi, good to see you again. I do want to start with AI in schools because your union and Microsoft have reached an agreement on safety and privacy standards on AI in schools. Can you just explain the background here and the talks and negotiations that you were having with Microsoft with Anthropic, with OpenAI, and give us the terms of this Microsoft deal?

Randi Weingarten (0:39)

Yes. So look, we started this AI Institute about a year ago, actually last July, with Microsoft, with Anthropic, with OpenAI. And because we wanted to actually learn the lessons from social media and from all of EdTech and schools, and have the teachers, not the technology companies, in charge of teaching and learning. I know I'm being overly simplistic, but frankly, Microsoft and others saw that as well, because they saw that over the course of the last 20 years, when teachers are not in a driver's seat, there's a lot of misfiring, and there's been a lot of misfiring in terms of EdTech. So we started that to help teachers learn, not to help kids learn using AI, but to help teachers. And what we said from the beginning, and Microsoft has been our lead partner in the tech world, was we know that we needed safety guardrails, not just, sorry, not my dog, not just best practices, but safety guardrails in terms of the tech stuff.

And we wanted to make common cause with parents about those kind of guardrails. And so it took a year, but we got there with Microsoft, and hopefully we will get there with OpenAI and with Anthropic, because we wanted to assure parents that children's data would never be sold, would not be used to test AI, would not be used to train, and we did a whole lot of other audit functions to assure that that would happen.

Carol Massar (2:39)

So what is the status of the discussions with...

SPEAKER_1 (2:40)

Sorry about the dog.

Carol Massar (2:41)

No, that's okay. We love dogs here. What is the status of your discussions with OpenAI and with Anthropic? Like when can we expect a deal from you and them?

Randi Weingarten (2:49)

You know, I hope shortly.

I think that because Microsoft has been our lead partner, it became important to just actually start very much focusing on one of the tech companies, but I think that we're going to see something shortly with OpenAI and with Anthropic. Our conversations with both have been going pretty well. I think that the real issue becomes what happens with Google and what happens with others, because this is part of the reason it has to be an industry standard. It can't just be, I mean, I'm very appreciative, and it took a long time to get there, but we kept saying to Microsoft, it can't just be about this little AI Institute. We have to actually make this real for parents and for educators and for school systems across the country. And the fact that they have committed to saying that this protocol, this safety protocol, is going to be part and parcel of every one of their contracts with school districts across the country November 1st, that is a really big goal, and that is a really big promise.

Tim Stenovec (4:10)

Congratulations. One of the things I want to ask you though, Randi, it always, what's the responsibility of companies that they shouldn't even have to do a pact or agreement, but understand the impact of the products they are creating, just out of the gate?

Randi Weingarten (4:26)

Look, I completely agree with you.

And part of this, and I don't mean to pass the buck, but part of this is, this is a whole new technology. Safety for the American people is as important as innovation. Innovation happens when people trust things. And so, the fact that safety has not been first and foremost in the virtual space is problematic. But that's why we thought that since all three of these tech companies want teachers to actually know what is going on with technology and use it properly, and they understand that teachers are the key to families being able to use this properly. I want to get to one other point about teaching and learning, that I'm glad we got there, but it took a year to get there. Yeah. It should be in rules. It should be in regulation. It should be updating FERPA.