Got a new development around Chinese AI startup Deepseek.

It says it's now temporarily restricting who can log into its chatbot app,

allowing only China-based mobile phone numbers because of large-scale malicious attacks.

American users can still use their email to sign in.

Adir Drabosa digs into the implications of its rise among U.S. consumers for today's tech check

ND. You were early, maybe first on this whole thing. We knew it was big maybe a month ago,

but we never imagined it would be this biggest story, even moving the markets this morning.

But this latest development, a potential cyber attack, it really adds to the intrigue surrounding

this Chinese AI lab that has really upended everything we thought we knew about the race in

AI and raising the stakes because it's already been unleashed here.

American users have flocked to the DeepSeek app en masse.

Over the weekend, I watched in disbelief as it climbed the app store charts, landing

at number one. Sometime yesterday, overtaking, guess who, chat GPT. Today it's blowing up on Google

trends. Now, consumer adoption marks a fundamental shift in the AI race. Until now, much of the

momentum has been developer-driven because this is an open-source

model. So that means that developers can use and tweak the model to suit their own needs, whether

that's removing censorship or building entirely new applications on top of it or using its infrastructure.

Still, that's something that perplexity CEO, Arvin's three-de-vas, told me could be concerning.

What's more dangerous is they have the best open source model, and all the American developers are building on that.

That's more dangerous, because then they get to own the mind share, the ecosystem.

If the entire American AI ecosystem, look, in general, it's known that once open source

is caught up or improved over closed source software, all developers migrate to that.

...