It’s no secret that spam and fraud phone calls are a massive problem. U.S. consumers receive roughly 4 billion robocalls each month, according to the Federal Communications Commission. Juniper Research estimated the worldwide cost of robocall fraud at $40 billion in 2022. While the rates of nuisance and fraud calls vary by country, they occur virtually everywhere.
What Is a Robocall?
A robocall is a computer-generated phone call, made with an automatic dialer, that issues a recorded response when the phone is answered. Not all robocalls are illegal, though some are scam calls and their intent is to deceive the recipient.
The problem is getting worse. Spam and fraud calls were up 30 percent in 2022, according to our latest testing. Imposters were the most reported type of robocall in FY2022, the second year running, according to the Federal Trade Commission’s annual report on the issue.
This is eroding trust in the voice call, the communications channel people prefer most for its speed, ease and personal touch. That’s a huge problem for mobile carriers and other businesses, as well as the consumers they serve.
Why Are Phone Scams So Prevalent?
Consumers receive so many spam and scam calls because they work and they’re profitable for the businesses that make them. Sophisticated spammers and scammers run just like real businesses, leveraging data and technology to constantly evolve and find new ways to maximize their bottom line.
The illegal caller industry in particular is comprehensive in its approach to developing new tactics and targets. (Not all spam calls are illegal; some unwanted calls are legal but may be the result of businesses not employing good calling practices.) It’s an organized industry, complete with its own KPIs and revenue targets.
Consumers receive so many spam and scam calls because they work and they’re profitable for the businesses that make them.
A limitless supply of low-cost computing power, plus increasingly sophisticated methods such as AI and other advanced technologies, are enabling bad actors to unleash new threats and evade traditional detection and defenses.
Solving the spam and fraud problem requires the same sophistication. You need AI to fight AI.
The bad actors have become too sophisticated and organized to be effectively mitigated by manual approaches. It’s not a puzzle that can be solved once; it’s a cat-and-mouse game that requires automated, human-level insight and intelligence to stay ahead.
4 Things AI Needs To Fight Spam and Fraud
In the fight against spam and scams, here’s the important litmus test: Your AI is only as good as its ability to adapt to change.
Imagine that you had a handful of data points and variables to work with. A group of robocallers ran various scam campaigns, and you tried to come up with a way to identify and stop them. You eventually came up with a list of known numbers or tactics they used for their campaigns and systematically blocked them. Problem solved.
That approach used to work, more or less, but no longer. Spammers and scammers realized that they had their own adversary — you — and they began seeking out ways to dodge and weave to avoid detection.
That’s where AI comes in. They use it. You need to, too. Here are four critical characteristics you need to win this fight.
Continuous and reliable data
AI fundamentally depends on a steady stream of high-quality data. You either need to ensure you can source, integrate and manage that data on your own. Better yet, you can enlist technology partners who focus on this problem 24-7.
One of the biggest misconceptions about AI is that it replaces people. In the fight against spam and fraud, this couldn’t be further from the truth. AI depends on people. The adversary here is fundamentally human. Good AI requires smart human oversight and data science, which means people and partners invested in understanding the illegal caller industry on a global scale, its strategies and how best to turn the knobs and levers of AI solutions to gain an advantage.
They also need a deep understanding of the regulatory environment and overall expertise in the voice industry so they can attack the illegal caller problem with complete context.
Whether it’s illegal callers or the legal nuisance, your AI needs visibility into the always-evolving mix of tactics and trends these callers deploy to ensure it is aggressively weeding out unwanted calls while ensuring desirable calls connect.
This requires significant visibility and research into their behaviors and signals, essentially a mix of #1 (data) and #2 (people.) You need to know what tools the spammers are using, the types of campaigns they are running, and the different tactics they are trying to boost their success rates. This is how we train AI on what to watch for in a constantly changing landscape. That’s how you reliably spot scammers and distinguish them from legitimate, wanted calls.
A source of truth
Each of the above needs a reliable source of truth, not guesswork. You’re essentially trying to train a machine to accomplish a goal, and that machine needs to know what the goal is. AI needs to be able to identify which calls are spam, and which are not. Then it can spot them based on whatever tactics and targeting they’re using at a given point in time.
This requires reliable sources of truth so AI can adapt as spammers and scammers tweak their tactics, including information from end users that can be fed back into the system. User reports at scale become that source of truth, augmented by a continuous stream of call engagement data in general, such as unanswered calls and quick hang-ups.
It’s Time To Fight Back
The bad news is that the spam and fraud problem isn’t going away. It’s getting worse.
The good news is that AI gives us the most powerful tool in this fight, one that can beat the bad actors at their own game. If you don’t invest in this fight, you’re only going to catch the less sophisticated spam. That’s a drop in the bucket, which means you’re going to lose the larger fight against sophisticated spammers who are eroding trust in voice calls.
The data tells us that scammers are heavily invested. It’s time to use AI to beat them at their own game.