AI scammers may be better at building trust than humans
Ars Technica AI
1 August 2026
How this applies to you
If your team handles payments, supplier changes, customer data, or account access, treat friendly messages as part of the risk. Add a second check for bank detail changes, urgent payment requests, password resets, and anything that asks staff to bypass the normal process.
Research covered by Ars Technica found that an AI chatbot was more effective than human scammers at creating exploitable trust. The concern is not just better phishing emails, but longer conversations that feel calm, patient, and convincing.
Why it matters. If your team handles payments, supplier changes, customer data, or account access, treat friendly messages as part of the risk. Add a second check for bank detail changes, urgent payment requests, password resets, and anything that asks staff to bypass the normal process.
