C:\>DIR Global ‘Agentic Regulator’ Hackathon

AI-Driven Fraud & Scams

Combat the industrial-scale use of AI for criminal purposes to create deepfakes, synthetic identities, and hyper-personalised frauds and scams in order to protect consumers, financial systems and society.

Context for the Problem Space

AI-enabled fraud and scams are rapidly accelerating, with agentic systems autonomously executing fraud and scam campaigns and creating deepfakes and synthetic identities at speed, bypassing traditional onboarding and corporate controls.

Key risks include but are not limited to:

  • Rules-based AML/CFT systems overwhelmed and unable to detect transactions linked to AI-enabled fraud and scams.​

  • A-KYC exploitation via AI-generated documents, face swaps, and deepfakes.​

  • Intelligence gap where authorities react weeks after fraud and scams have generated loses​.

  • Industrial-scale fraud and scam operations including polymorphic phishing intersecting with organized crime and human trafficking.​

  • Erosion of public trust through deepfakes impersonating national authorities and public officials.

Potential Solution Areas

Adapting Beyond Static Rules

How can defenders detect AI-enabled  fraud and scams when adversaries use AI to bypass static thresholds and change their tactics daily?

Securing Identity Verification

How can institutions protect digital onboarding from AI deepfakes and forged documents that easily bypass standard KYC controls?

Countering AI Social Engineering

How can authorities disrupt hyper-personalized, real-time AI frauds and scams that exploit digital footprints to bypass traditional security safety nets?

Combating Industrial Fraud & Scam Networks

How can national authorities and enforcement agencies scale cross-border cooperation to match the speed, volume, and transnational nature of automated scam operations?

Restoring Trust and Authenticity

How can authorities verify official communications and protect the public from fraud and scams that undermine genuine warnings and evidence and support victim redress?

Illustrative Hackathon Prototypes

‍Scam Typology Radar

A monitoring swarm that scans open and dark web channels to detect emerging deepfakes and scam campaigns, mapping tactics to MITRE ATLAS to alert authorities early.

Typology-to-Rulebook Copilot

An analysis agent that mines anonymised incident reports to pinpoint exactly where current rules and red-flag guidance lag behind live criminal practice.

Triage and Disruption Agent

An FIU and law enforcement copilot that clusters incoming reports into criminal networks using graph analytics, prioritizing cases by threat level and triggering takedown activity.

Federated Fraud Intelligence Exchange

A secure network for cross-border authorities to share structured fraud and scam typologies, signals intelligence, deepfake signatures, etc. via privacy-preserving federated models and communications channels without exposing raw personal data.

Adversarial 'Honeybots’

Active intelligence agents that pose as vulnerable victims to infiltrate live scam networks, extracting operational data - like active mule accounts - to stop fraud and scams in real time.

Reference & Resources

Contact

For additional questions, please contact: contact@cdir.global