Updated
Updated · Fortune · Aug 1
Arup Employee Sends $25.6 Million in Deepfake Scam After Fake Executive Video Call
Updated
Updated · Fortune · Aug 1

Arup Employee Sends $25.6 Million in Deepfake Scam After Fake Executive Video Call

3 articles · Updated · Fortune · Aug 1

Summary

  • $25.6 million was transferred by an Arup finance employee after a video call featuring synthetic versions of senior executives delivered false payment instructions.
  • The case is presented as part of a wider trust problem in AI, where realistic voices, faces and outputs can accelerate fraud and errors before workers detect them.
  • Starbucks scrapped an AI inventory system after nine months because baristas said it miscounted products and slowed work, while Deloitte Australia partially refunded a $290,000 AI-assisted government report containing fabricated sources.
  • The commentary argues companies must build verifiable identity, traceable data, independent testing and clear accountability into AI, warning that innovation without trust stalls and raises operating costs.
  • It frames that challenge as strategic as well as commercial, saying the AI ecosystem that earns the most confidence will attract more capital, partners and adoption.

Insights

Could the AI tools designed to boost your company's productivity secretly be the ultimate trojan horse for corporate sabotage?
If seeing and hearing are no longer proof of reality, how will your company prevent the next multi-million dollar deepfake heist?
When digital identities can be perfectly cloned, is returning to analog, in-person verification the only way to secure high-stakes deals?