AI and the New Era of Fraud: How Authoritative Telephone Intelligence Shields Business Communications 

By Steve Tang, Chief Technology Officer, iconectiv

Businesses have always had to deal with scams, but the sophistication and scale of this new era of exploits are beyond anything of the past. Thanks to the power of artificial intelligence (AI), well-funded and sometimes nationally supported networks of fraudsters devise schemes that result in significant financial losses for companies, not to mention lost customer trust and compliance issues for those operating in regulated industries. 

These AI-enabled scams can be especially overwhelming for IT departments, which must continue to dedicate more resources and manpower to detecting, preventing and resolving these attacks. Unfortunately, many fraud detection systems are known for producing false positives, resulting in alert fatigue among personnel. To make matters worse, plenty of IT teams are understaffed, with research from Deloitte finding that almost 90% of IT leaders say that recruiting and retaining tech talent is an ongoing challenge. 

The good news is that AI isn’t reserved for the bad guys – it can also be leveraged by legitimate businesses looking to enhance their fraud detection methods in order to protect their revenue, brand reputation and bolster customer trust. One way that businesses can effectively combat AI-driven communications fraud is by using authoritative telephone numbering intelligence to identify and block fraudulent voice and short message service (SMS) communications proactively.

How Scammers Use AI: Deepfake Voice Clones

Before discussing how businesses can stop schemes, it’s important to examine what types of AI-powered fraud they are up against, specifically those that use voice and SMS messages, which are a favorite attack vector of cybercriminals today. In the past, wary customers could recognize when a scammer attempted to impersonate a government agency, major brand or known business over the phone or SMS. Today, various AI technologies allow fraudsters to create deepfake voice clones for robocall scams that are much trickier to figure out.

Unlike the past scams, which sounded more robotic than human, this new generation of scams will use convincing AI-generated voice clones of real people familiar to the victim to trick them into divulging sensitive account information or sending money. There are several ways scammers can generate these voice clones, such as getting a voice print from a voicemail greeting or a video on social media, or even calling a target and recording the conversation. 

Fraudsters, using the voice of a friend or family member, will call victims and insist they need to give them money. Insidiously, fraudsters will also use AI to copy the voice of a victim’s child and make them sound distressed. Posing as kidnappers, the scammers call the parents and demand a ransom for the release of the child. Additionally, bad actors use AI voice clones to masquerade as someone’s manager or boss and request that the victim withdraw and transfer funds to pay for a bogus business-related expense. 

For IT departments, these scams can be a nightmare to address without the proper telephone numbering intelligence. With AI-powered schemes considerably more challenging to perceive, scammers are finding greater success at the expense of legitimate businesses and the general public. A 2023 survey by Regula discovered that 37% of organizations experienced deepfake voice fraud. A comprehensive study of global call fraud found that the average American fraud call victim lost $539 in 2024. 

Authoritative Phone Numbering Intelligence

To overcome these communication schemes in which bad actors use and abuse communications networks, businesses can leverage telephone numbering intelligence to train their AI solutions to differentiate between normal and fraudulent activities. Of course, verifying someone’s digital identity is difficult without a trusted and authoritative source. Consider how generative AI models are only as accurate or helpful as their training data. This same principle applies to fraud prevention. The more high-quality data one feeds an AI model, the more effective it is at distinguishing a legitimate caller from a scammer. 

A comprehensive, authoritative database of verified ownership information serves as a crucial reference source when paired with AI analysis tools. By enabling AI systems to query such a database during verification processes, organizations can better assess the likelihood that a caller or texter is who they claim to be, supporting efforts to safeguard revenue and brand reputation while boosting customer confidence in communications. 

Should the AI system identify potential inconsistencies or risk factors that the user of a telephone number is potentially not who they say they are (making them a possible fraudster), it can require that this person provide additional information, such as multi-factor authentication or verification through established secure channels. 

It is also worth noting that authoritative and deterministic telephone numbering intelligence data is crucial for thwarting other popular schemes like SIM swaps and port outs. Here, AI models, having access to an independent resource detailing information for each telephone number, can determine when a phone number was ported or if it is associated with a particular SIM, line type or location. 

Why Protecting Phone Communication is so Important

Phone calls and SMS messages represent invaluable channels for conducting business and engaging with customers, enabling everything from appointment reminders to product promotions. Research shows that phone calls remain the preferred means for consumers to interact with businesses despite the availability of other options like social media. 

As such, businesses must safeguard the integrity of mobile phones against sophisticated AI-generated deepfake voice clones. By pairing a comprehensive, authoritative database of verified ownership information with AI analysis tools, enterprises can effectively prevent the latest fraud schemes and restore customer trust in business communications, ultimately protecting revenue and brand reputation. 

Steve Tang is Executive Vice President, Chief Technology Officer and Head of Engineering for iconectiv. He is responsible for consolidated software development, quality assurance, system/usability engineering and supporting business partners with emerging technologies.

Tang is a seasoned professional with more than 20 years of experience developing highly scalable, robust products that enable the seamless interconnection of devices, applications and networks globally. He previously held positions at leading telecommunications companies, including Motorola.

Tang holds a Bachelor of Science degree in Computer Science from Rutgers University.