GEOFF BENNETT: Now to our series on the future of A.I., one of the biggest concerns, how children and teens use it daily.

Companies are taking some steps.

Meta, for example, announced its A.I.

tools will now notify parents if its A.I.

systems believe a teen may be considering suicide or self-harm.

But many say the companies are not doing enough, can't be trusted, and argue more regulation is needed.

The tech philanthropy the Omidyar Network is out with a new approach of its own.

Michele Jawando is the CEO.

And I spoke with her yesterday.

Michele Jawando, welcome to the "News Hour."

MICHELE JAWANDO, CEO, Omidyar Network: Thanks so much for having me.

GEOFF BENNETT: So much of the debate around kids and A.I.

has focused on whether kids should be able to access it entirely and at what age.

I know you say that's the wrong question.

Why?

MICHELE JAWANDO: Yes, I think -- listen, I think access and age verification are a part of the conversation, but we shouldn't end the debate there.

The onus of putting kind of restrictions again falls on families.

It falls on parents who are just trying to figure it out.

What we really want to ask are companies to design safer products.

And, at the end of the day, you have the world's best engineers, who know how to do this, versus parents, who are just trying to figure it out.

And so for us at Omidyar Network, we really focus on design and design standards as a means to a more safe and healthier ecosystem.

GEOFF BENNETT: Nearly nine in 10 kids and teens are already using A.I.

That's according to a national survey.

So, to your point, if banning it isn't really realistic, what protections should exist?

What should those guardrails look like?

MICHELE JAWANDO: So there are a few things.

You have said it so eloquently.

Nine in 10 kids are already using A.I.

Our partners at Common Sense Media share that young people, about a quarter of young people are using A.I.

every day for something.

So the question has to be then on the companies and the platforms themselves to develop safer.

One, how do you restrict manipulative and addictive practices?

That's something that we all can agree on, particularly things like romantic attachment.

The second thing that I would then point to are safety designs.

So how do you have built-in brakes?

How do you make sure that we're not capturing data from young people who are just going on different apps?

And many of the apps that young people use, three and four actually take, share, and sell their data.

So how do you provide greater design features that protect our young people?

GEOFF BENNETT: The House recently passed the KIDS Act, the KIDS Safety Act.

This regulates social media, video games, A.I.

chatbots.

Does that legislation meaningfully address the range of risks as you see them?

MICHELE JAWANDO: The KIDS Act -- and I'm very happy to see Congress start moving in the direction.

Eighty percent of Americans have said we want to see more in this area.

But, again, I think it is the beginning of the conversation.

I don't think it is anywhere near the end.

And there are a few key reasons.

Again, we're putting the onus back on parents.

If most of the work in the KIDS Act is really focused on parental notification, again, I'm a parent.

I'm trying to figure it out.

I'm a parent of four.

What am I doing here?

How am I supposed to lead on this, versus companies who know and have the technical standards and expertise?

So that's the first thing.

The other thing is they're missing something that's called the duty of care in the recent bill that was passed.

And that's a legal term, but essentially what that says is, as a company, if I'm designing a product, I have to design it as safe as I possibly can.

That's all we're saying.

And that's missing from the current bill.

GEOFF BENNETT: So walk us through how a chatbot, an A.I.

chatbot, should behave when it knows it's speaking to a 13-year-old.

MICHELE JAWANDO: Well, first off, I think we believe at Omidyar Network that these are design standards that should apply to everyone.

We actually just launched a partnership with AARP, and we're seeing some of the same nefarious design standards really attacking older Americans, who are also in a loneliness crisis.

So I think design standards that don't focus on manipulating behavior, particularly romantic attachment, or design standards that say here's a brake notification, should be kind of commonsense, basic.

We also want to get away from taking people's data as part of this compromise for using chatbots.

And, finally, having a real conversation with these companies, who know what they can do here and know how to create a safer environment is something that they can and should do, not just the individual consumers who are trying to figure this all out.

GEOFF BENNETT: Your point about data is a good one because there is this concern, how do you actually verify a child's age without creating new privacy risks?

MICHELE JAWANDO: That's right.

So one of the things that we're seeing in -- here in the U.S.

and in countries around the world is the advent of kind of age verification.

But, again, we don't really, particularly here in the U.S., have a privacy kind of safeguard for all of our information.

And so often what's happening is, we're asking companies, private companies, to share my personal I.D.

And yet we don't have any privacy regulation around that.

So we actually believe that there are other ways to address that issue.

One of the things that we have seen are age estimation.

We also, frankly, believe that, if you have privacy baked into your application from the beginning, those are things that you don't necessarily have to have.

GEOFF BENNETT: You know, we're seeing schools restrict smartphones, which is something that most parents support.

Some parents might ask, why should A.I.

be treated any differently?

MICHELE JAWANDO: Yes.

Look, I think the bell to bell -- and we're seeing this in schools across the nation, that they have a ban on smartphones.

And I think that, again, is a part of the conversation.

But access is just one step.

If you and I step back as parents and we really think about this, I can restrict some of the behaviors, but just because I take that away doesn't mean that that application isn't doing something nefarious when they actually pick it back up at 5:00.

And so my question is, are we as a society ready to say we want a better designed product, point blank?

It doesn't matter when you pick it up, whether it's before school or after, but we want a product that's safe for our young people.

And I think we can do that.

GEOFF BENNETT: When it comes to higher education, we saw this striking example at Brown University.

GEOFF BENNETT: And you're laughing because you know where I'm going with this.

MICHELE JAWANDO: I am.

GEOFF BENNETT: So scores on an online midterm, they rose dramatically, and then they dropped when the final was in person, presumably because students had less access to A.I.

Does education itself need to be redesigned to address the reality of artificial intelligence?

MICHELE JAWANDO: There is a fundamental question that I think we have to have in society right now and be prepared to answer.

For the last 20 years, we focused on, how do we develop a knowledge economy?

And then we have created perhaps one of the greatest inventions to help us really think about the vastness of knowledge and information, and it is easily at our fingertips every single day.

That is a great thing.

And yet what we haven't really reckoned with society is what that then means for the process and the systems of how we have thought about what education should do.

In some ways, I understand those young people.

They are looking at a job market that is constantly -- they're constantly hearing A.I.

will replace them, so they have to be at the top of the class.

We're also hearing every day that we're in the a global race around A.I.

So you have all of these pressures, and then we say you have to know how to use A.I.

You have to be A.I.-literate.

And then we have education system that hasn't really caught up and decided how and -- how we use education in the classroom, but yet everyone has it as a part of their curriculum.

So these are real questions about how we learn, what we learn.

I don't think we can ask students just to give rote information or responses.

There is a new way of learning that I think will require all of us to think about critical thinking, about how we grade, about how we talk about these questions.

GEOFF BENNETT: Absolutely.

Michele Jawando, CEO of the Omidyar Network, always a pleasure to speak with you.

MICHELE JAWANDO: Oh, thanks so much, Geoff.