RED-AID : REspectful and capability-centreD integrated AI Device for preventing call fraud
New method for identifying and preventing phone fraud, which uses recent advances in neural natural language processing, machine learning and artificial intelligence.
Applications: Call fraud detection and prevention
| Features | Benefits |
|---|---|
| Method for determining the likelihood that part of a call or a whole call is fraudulent | Protection and prevention for users, vulnerable or not, from different types of fraudulent calls |
| As a first step the method uses a neural network to transcribe the content of the exchange | The novel method can identify different types of fraudulent activities such as transferring money to a bank account or purchasing a certain product |
| A natural language processor is trained using artificially synthesised conversations. It then identifies and categorises the content of the exchange | The method guarantees reliability of fraudulent call detection, even when the fraudulent transmitter changes tactics over time to avoid detection |
| Using a knowledge-based risk assessor, the method then determines the likelihood that the call is fraudulent | The knowledge-based risk assessor uses information collected during previous calls and develops a set of evidence and rationale before determining if the call is fraudulent |
| If a call is determined as being fraudulent, it can, in certain situations, be terminated without any input from the receiver | The automatic termination of the call, if fraudulent activity is detected, protects users with an increased vulnerability to fraud (i.e. users with cognitive impairment) |
Patented and Available For
- Co-development
- Consulting
- Licensing