Arnab and Dev, founders of a pig-butchering scam operation running out of a multi-story building in south Calcutta, were reviewing their monthly numbers.

“Revenue is flat,” Arnab said, tossing the ledger onto the desk. “Forty-two staff, and every one needs salary, chai, and supervision.”

Dev leaned back. “Our scammers can only handle so many targets a day. That’s our ceiling. Last week we lost three promising targets because nobody answered in time.”

“Then we raise the ceiling.”

“How?”

Arnab tapped his laptop. “Muse. Meta’s AI agent everyone’s talking about. It never sleeps, never gets bored, never forgets a target’s name.”

“You want a machine running our scam?”

“Muse can send thousands of messages at once, in flawless English, across every time zone. We keep a few people for the final stage, the money transfers. The rest we let go.”

“Romance, crypto, fake investments,” Arnab added. “Muse can run all of it.”

“And the targets won’t suspect a thing?”

“Why would they? Muse can mimic our best scammers. It can tailor its response based on what the target is saying at the other end.”

Dev counted on his fingers. “We can let go of thirty scammers. That’s seventy percent of the payroll.”

“And Muse reaches sixty percent more targets than we do now.”

Dev scribbled numbers, then stopped. “Profits up eighty percent.”

“Eighty,” Arnab said, grinning. “And nobody demanding overtime.”

Dev poured two glasses of whisky and raised his. “To the new business model.”

“To the new business model.”

They drank. Dev looked up, smiled, and said, “Thank you, Muse.”

Two months later, Arnab stormed into Dev’s office, slamming a report down. “Targets up sixty percent. Staff down seventy percent. But you know what? Profits, instead of going up by eighty percent, are down by eighty percent!”

Dev stared at the report. He turned to his computer. “Let’s pull up the chat messages between Muse and our targets.”

“We trained Muse to select the targets carefully,” Arnab said. “Old. Lonely. Vulnerable. Some would say no; we knew that. But the majority would pay up. That’s the whole business. So why isn’t it happening?”

Dev scrolled the chat history. “Look at this. Nobody says no. ‘Let me get back.’ Then silence. Then ‘Next week, I promise.’ Hundreds of them.”

“Wait.” Arnab leaned in. “I’ve seen this line before. ‘So sorry, busy week.’ Here. And here. And here.” He grabbed a notepad and tallied. His pen moved faster, then stopped. “Eight. Just eight excuses. Bank delay, busy week, waiting on a transfer, traveling, family emergency, a wedding, tax season, been sick. Every target’s response can be tied to one of these excuses.”

“That’s impossible.”

“Then explain this. Ohio, Texas, Oregon, Florida, Michigan. Different states, different backgrounds, yet they all write the same way. Same warm opening, same little apology, same exclamation mark.”

Dev didn’t answer. He took a breath and stared at the screen. Hundreds of eager, patient, endlessly apologetic replies scrolled past, every one politely promising to pay next week.

FBI Special Agent Hale had just finished presenting the scam statistics when his supervisor Ruiz said, “Scam complaints are down by ninety percent. The dollars lost to the scammers are down even more. These are incredible numbers!”

“Thank you, sir.”

“And you did all of this with just a two-person team?”

“Two people and an AI agent. The one that kept saying, ‘I’ll pay you next week.’”

“Oh, yes,” smiled Ruiz. He looked up and said, “Thank you, Muse.”

Check out my other story:

A Walk Through Times Square

Thank you, Muse was originally published in Microcosm on Medium, where people are continuing the conversation by highlighting and responding to this story.