Newspaper's 37th Anniversary Special Project AI Great Transformation, Finding Direction ③ Safety/Rights AI Threatens My Information and Wallet

"Mom, my phone screen broke and I need to send money urgently for repairs."

The trembling voice of a child heard through a messenger or phone line. Voice phishing that used to end with awkwardly worded text or crude composite photos in the past has evolved into 'super-intelligent crime' when combined with generative AI. With just a few seconds of SNS video or audio file, deepfake technology that perfectly reproduces the intonation, breathing, and even facial expressions of a voice has come to the forefront of fraud methods.

Artificial intelligence, disguised under the mask of convenience, has now emerged as the most covert and powerful threat to consumers' assets, information, and rights.

"It really sounds like my child's voice"…The terror of deepfake phishing raised by AI

The patterns of phishing damage reported to financial authorities, police, and the global security industry are of a different dimension from the past. As AI deep voice technology becomes more widely available, variant scams involving replicating the voices of family members or acquaintances to demand money are rapidly increasing.

According to a recent survey by global cybersecurity company McAfee, more than 10% of adults worldwide have either directly experienced AI-enabled voice cloning fraud or witnessed damage to those around them. Overseas, there have been major incidents where deepfake technology was used to precisely synthesize video call screens of acquaintances to extort hundreds of millions of won, and domestically, reports of super-intelligent deep voice phishing targeting parents' anxiety continue to mount.

In the past, crimes could be suspected through typos or awkward speech patterns, but now AI analyzes the target's SNS traces and implements customized scenarios that reflect their usual speech patterns, personal relationships, and even personal information, making it virtually impossible for ordinary consumers to instinctively distinguish them.

The same product, but it costs more for me?…'Price discrimination' created by AI

Another major culprit quietly depleting consumers' wallets is the rampant spread of 'dynamic pricing' utilizing AI algorithms.

Variable pricing, which was mainly used for airline tickets and accommodation reservations, is now penetrating daily consumer spending across e-commerce, daily necessities, and mobility services as it combines with AI's hyper-personalization technology. AI analyzes consumers' access devices, past purchase histories, shopping cart abandonment time, and even real-time location data. Based on the analyzed purchasing patterns, it presents prices higher than the regular price to consumers who are likely to buy right away, while displaying discount coupons only to hesitant consumers.

Policy authorities such as the U.S. White House Council of Economic Advisers (CEA) have also recently warned through reports that algorithmic price discrimination impairs consumer welfare and deepens information asymmetry. While it appears to reduce the fatigue of searching for the lowest price for consumers, it turns out that they are exposed to structural price discrimination where they pay higher costs than others by being deceived by big data algorithms.

"Once breached, tens of millions of people"…Massive personal information leaks in the AI era

As AI models become more sophisticated, the volume of personal information collected to train and operate them is growing exponentially. The problem is that if AI systems or conversational services are hacked, the scale of damage becomes incomparably larger than in the past.

In fact, global market research firm Gartner predicts that as AI adoption accelerates, privacy breaches and data leak incidents will establish themselves as one of the most critical risk factors for enterprises. There are also frequent reports of security vulnerabilities where corporate confidential information or personal financial information is carelessly entered into conversational AI search services, only to have the data leaked externally or exploited as training data.

Professor Yoshua Bengio, a pioneer in artificial intelligence research, pointed out that "an AI system lacking control could be a potential disaster that shakes society as a whole in terms of security and ethics," strongly urging the establishment of safety measures for the technology.

Behind the convenience of technology, protecting 'consumer sovereignty'

Experts warn that if the safety and rights infringement elements brought by AI are left unaddressed, trust in the entire digital ecosystem could collapse.

If companies use AI algorithms to deceive consumers or arbitrarily discriminate in pricing, establishing institutional mechanisms to clearly regulate and enforce transparency is urgent. Additionally, while strengthening corporate responsibility for personal information protection, consumers themselves must develop 'digital defense capabilities' by thoroughly re-verifying suspicious voice and video calls and carefully reviewing personal information sharing agreements.

Confronting the dangers lurking behind the convenience brought by AI and building a safety net in which humans maintain control of technology is a task that can no longer be postponed.