tech
Why Incode is working to build defenses against agentic fraud
Agentic fraud, cons that machines plan and run entirely on their own, is sharply rising. Incode is working to build defenses against it.
TL;DR
- A new internet ritual involves asking suspected bots to write a poem about tangerines to identify them.
- Agentic fraud refers to cons planned and executed entirely by machines, targeting thousands of victims simultaneously.
- Artificial intelligence has removed humans from the fraud equation, enabling scams at scale without a human bottleneck.
- Generative AI makes scams cheaper to write, while agentic AI runs the operations, leading to incidents like data breaches and mass-produced romance-bait scripts.
- The cost and scalability of fraud are now limited by compute and budget, not human labor, with 'fraud-as-a-service' kits becoming available.
- Incode predicts agentic fraud will account for the majority of global fraud losses within two years.
- Traditional defenses like familiar voices or known faces can now be forged, making heightened caution insufficient; verification of a present human is key.
- The article suggests a future where individuals must assume that many online interactions are with machines designed to defraud them.