AI Voicebots Threaten Consumer Trust in Voice - How Do We Fight Back?

June 2024

Telecoms & Connectivity

Georgia Allen

Research Analyst

Fraud in the telecommunications market has existed for decades. However, with emerging technologies such as AI (Artificial Intelligence), the nature and significance of the risks posed by telecom fraud are greater than ever. More specifically, AI has added major complexities to robocall fraud due to the increased reality of the interactions; making it harder to distinguish between a legitimate and fraudulent call.

Robocall fraud means automated phone calls by fraudulent players, made to deceive mobile subscribers into giving away personal information such as bank details.

This type of fraudulent activity impacts a range of stakeholders:

• Mobile Subscribers: These are the stakeholders who are directly impacted by robocall fraud, whether through monetary loss or emotional impact. 
• Mobile Network Operators: If they allow a significant amount of fraudulent activity to occur on their network, they will be impacted by loss of trust. Enterprises and mobile subscribers will look to alternative providers if they feel at risk of fraud.
• Enterprises: If their enterprise's number is being spoofed, they are impacted by a loss of brand trust, and fraudsters can directly target them; causing monetary loss. 

Mitigation efforts have seen the introduction of STIR/SHAKEN by the FCC, and other attempts by other regulatory bodies to verify legitimate calls. However, technological advancements have disrupted the effectiveness of such attempts at regulating robocall fraud. 

AI ~ Fighting & Enabling Fraud

Previously, fraudulent players relied on making random calls on a large scale when attempting to defraud mobile subscribers. However, the availability of AI has presented a valuable tool to fraudsters. In particular, they have used AI voice cloning capabilities to enhance the reality of the voice on the calls. 

There is a cost to the fraudulent players when implementing AI into fraudulent activity, and as a result, the barrier for entry into fraud has been raised. 
 
The use of AI voice cloning by fraudsters is shifting the nature of robocall fraud away from mass random calls to more targeted attacks on mobile subscribers. Fraudsters select victims based on vulnerabilities, and exploit these to extract money from the individual.

This is particularly applicable if the voice being cloned is that of a family member of the mobile subscriber, due to the inability of the mobile subscriber to check that the person who is calling them is who they assume it to be. This advancement will make mobile subscribers doubt the caller’s identity on calls. 

According to our latest robocall mitigation research, in 2023, mobile subscribers lost approximately $65 billion to fraudulent robocalls. Over 50% of this global figure was accounted for by North America. However, we are seeing increased fraudulent losses to robocalling across all regions. 

This shows the prevalence of robocall fraud and, hence, the need for robocall mitigation stakeholders to act quickly to protect the integrity of the voice channel.

Is There Still Value in Voice?

Is the voice channel still needed with the rise in alternative communications channels such as OTT and SMS? 

The voice channel remains an important way to communicate for several use cases. For example, the healthcare industry heavily relies on this channel for communications between patients and healthcare providers. Conversations are usually sensitive and less suited to alternative communications channels. 

Therefore, voice is highly valued, meaning that the loss of trust in the voice channel resulting from robocall fraud will make communicating via voice channels more complex.

What Can Stakeholders Do?

International disparities in phone call regulations make establishing a comprehensive robocall mitigation strategy difficult. Nonetheless, robocall mitigation vendors, mobile operators, and regulatory bodies must take a number of important steps to re-instil trust in the phone channel. 

• Stakeholders must collaborate to establish and implement voice channel regulations. This will strengthen efforts by regulatory frameworks existing in line with solutions offered by robocall mitigation vendors to tackle robocall fraud.
• Regulatory bodies must work together with international regulators. Regulatory bodies need to work across international borders to regulate phone calls at their inception, targeting the countries in which large numbers of fraudulent players originate. 
• Re-instilling trust in legitimate calls is crucial. This element will improve enterprise productivity and help customers recognise a legitimate call.
• Enterprises must offer omnichannel engagement. The ability of enterprises to engage with their customers on a range of platforms will help them meet customers' preferences. Additionally, it will enable enterprises to pre-warn customers to expect a phone call by sending an SMS first; helping the customer identify legitimate calls and thus answer them.

Georgia is a Research Analyst in Juniper Research’s Telecomc & Connectivity team, providing analysis on the emerging trends and latest developments in the telecommunications market. Her recent reports include Retail Roaming, and Robocall Mitigation & Branded Calling.