Chapter 1

I first noticed something was wrong with Ayla during our third video call.

The image was perfect. Every single one of her eyelashes was there. When she pursed her lips, that slightly asymmetrical curve at the corner of her mouth was exactly how I remembered it.

What was wrong was her left ear.

That day, she was wearing a silver earring. The light was coming from the right side. The earring should have cast a shadow to the left on her earlobe.

But that shadow―it was on the right.

A real person's earring wouldn't cast a shadow in the opposite direction.

Unless―this was an AI-generated deepfake...

1.

I met Ayla during an online session of an AI founders' forum.

The moderator introduced her as an "Independent AI Ethics Researcher." Her profile picture was a black-and-white side profile―high cheekbones, short hair, and a very specific kind of stillness in her eyes.

I sent her a private message.

"What's your take on the ethical dilemmas of generative AI?"

She replied quickly.

"The biggest dilemma isn't that AI will lie to us. It's that after it learns how to lie, humans find it even easier to lie to themselves."

I read that sentence three times.

Then, I hit follow.

We started chatting. We talked about everything from AI to movies, from movies to Jorge Luis Borges, and from Borges to the seashells she used to collect on the Maine coast as a kid.

She said she lived in San Francisco. I told her I was there too.

"Then we're close," I said.

"Yeah."

"Want to grab a coffee sometime?"

She sent a smiley face.

"I don't drink coffee. Only warm water."

Someone saying they don't drink coffee and only drink warm water―that detail was too specific. It felt too grounded to be something an AI would just make up.

I asked her for a video call.

She hesitated for two days, then finally agreed.