Could that little TM protect against big AI?

From American Public Media, this is Marketplace Tech.

I'm Stephanie Hughes.

It's Friday, time for Marketplace Tech, where we highlight a few stories you may have missed.

This week, Taylor Swift trademarks her voice and image.

Plus, a Japanese airline experiments with humanoid robots to help with baggage.

But first, Google lands a deal with the Pentagon to let its AI models be used for classified

work.

A Google spokesperson told us in a statement that the company is proud to be providing its

services for national security and that it believes AI should not be used for, quote,

domestic mass surveillance or autonomous weaponry without appropriate human oversight.

I talked with Maria Curia at Axios about how Google's

deal compares to ones the DOD has reached with other AI companies. So Google's deal is being

described as more permissive because it is being compared to OpenAI's deal. And in the OpenAI deal,

the company essentially has said that it has full discretion over

its safety stack, whereas in the Google deal, they are going to be collaborating with

the Pentagon when it comes to those safety mechanisms.

And so that difference in language is what's leading people to think that Google's deal is not as strong.

But at the end of the day, Stephanie, all of these contracts, I think, are experts actually say are subject to potential loopholes and workarounds because the law has not actually caught up with

artificial intelligence. And you have folks over on Capitol Hill trying to figure out how to

modernize those laws as we speak. Yeah, you've been reporting on this. You've been reporting

on whether these deals could actually spur Congress to create new rules. How likely do you think that is?

...