If You Thought "Cheap"... You Have Donated Your Personal Information [AI Error Note]

Why Car Insurance Companies Offer Step Count Discounts
Predicting Other Risks Through Health Information
The Danger of the “Data Is the New Oil” Ideology

Editor's NoteExamining failures is the shortcut to success. 'AI Wrong Answer Notes' explores failure cases related to AI products, services, companies, and individuals.

I received a notification to renew my car insurance premium. I designed the insurance at the same level as every year. I took advantage of various discounts such as safe driving score discount, black box installed vehicle discount, and email invoice discount. While trying hard to save even a penny, a new discount item caught my eye. It was an item that had no discount benefit last year. It was the ‘step count discount.’

Why car insurance companies offer step count discounts

The back view of a person walking. Getty Images Bank

As soon as I saw it, the thought “Why would a car insurance company give a step count discount?” popped up reflexively. But then I realized, “Ah, it’s my health information value.”

Some domestic car insurance companies offer discounts on premiums based on the subscriber’s step count. The collection of step count data, which seems unrelated to car insurance, contains strategic intentions of the insurance companies.

First, based on the hypothesis that people who exercise regularly tend to have safer driving habits, this data can be used to assess the driver’s risk level. People who walk often or try to walk frequently likely have a strong desire to be healthy. Naturally, they would have a strong risk-avoidance tendency. Therefore, they are more likely to have a lower probability of traffic accidents. Insurance companies gain a competitive advantage over rivals in calculating appropriate premiums in the future.

Second, the step data collected by the insurance company serves as a foundation for cross-selling health insurance or life insurance products in the future. There is no insurance company that sells only car insurance products. They will be able to recommend various services and products based on everyday health indicators like step count.

Besides economic benefits, there is also the side effect of image improvement. The image of “a company that cares about the customer’s health.” Ultimately, the collection of step count data can be seen as part of a comprehensive strategy to enhance the competitiveness of insurance companies.

As digital technology deeply permeates our daily lives, all our actions are recorded. These records are called 'data.' It logs when, where, what, and how something was done. Everything?what we bought, what we watched, what we did?is stored somewhere on invisible servers. Pixabay

“Data is the new oil.”

This phrase has been circulating for quite some time, but it has now become the dominant perspective on data. My step count data was like oil to the insurance company. Although it was an extremely tiny amount, not even a single drop.

Oil gains value through refining and processing. The collection and utilization of step count data by insurance companies can be compared to the petroleum industry. Raw data in the form of step counts is gradually refined through lifestyle pattern analysis and risk assessment. This refined data is reborn as new insurance products or customized services, creating much higher added value.

The war to secure data as the new oil is not limited to the insurance industry. Supermarkets, online shopping malls, Facebook, Instagram, Google search, hospitals, and even governments are involved.

Everyone is desperate to secure data. Under the pretext that “AI technology advances through more data, which leads to better services,” they argue that vast amounts of data are necessary to improve the safety of autonomous vehicles, enhance the accuracy of medical AI, and provide personalized services. Every move of individuals is subject to collection and analysis. Step counts, movement routes, consumption patterns, SNS activities ? all our actions are treated like fuel for AI development.

In recent years, a consensus on personal information protection has formed and regulations are gradually being established. The appearance of obtaining consent for personal data collection is also taking shape. However, it is hidden in dozens of pages of terms and conditions in very small print or buried in the tedious repetition of signatures. In the name of corporate and national interests, the value of “my data, my personal information” is very small and humble.

The misconception that data is a natural resource and that “first come, first served” is everything

The appearance of an oil drilling rig. Getty Images Bank

However, the phrase “Data is the new oil” should not be taken as an absolute truth in the AI revolution era. The infinite expansion of this logic can lead to paralysis of ethics and morality. Consider the following cases.

At the University of Colorado Springs in the United States, a professor installed cameras in the middle of the campus. The faces of about 1,700 students, faculty, staff, and citizens entering and leaving the campus were secretly photographed. Between 2012 and 2013, over 10,000 photos were collected, and this data was used solely to train the facial recognition system the professor was developing.

There is also the case of Microsoft’s (MS) “MS-Celeb.” MS-Celeb was the world’s largest facial recognition dataset in 2016, with about 10 million face images collected. Where were they collected from? They indiscriminately scraped photos of 100,000 celebrities from the internet. There was no prior consent or notification.

Once you learn about some of the celebrities included in this dataset, you realize how “recklessly” data collection was conducted. Photos of human rights activists and writers who usually emphasize the importance of personal information protection and worry about privacy violations by facial recognition systems were also collected. Literally, it was a reckless collection, and an ironic case.

'The implication of the phrase “Data is oil”'

Smartphones are a source from which data springs forth every moment. Pixabay

At this moment when the logic of indiscriminate data collection dominates the world, we are missing something important. Our data records are not just simple numbers or records. They are footprints containing individual lives and experiences. Someone’s location information contains many relationships and life routes of that person. Consumption data reflects personal preferences, economic status, and lifestyle.

Data is not a colorless, odorless abstract concept but vivid pieces of individuals. When personal daily life is reduced to infrastructure for AI development or raw material for industry, the human context and meaning contained within disappear.

Of course, we cannot go against the flow of the AI revolution. The importance of data will only increase. However, this should not be done by ignoring the value of personal information or deceiving individuals. What we need is a balance between technological advancement and human values. We can demand a new paradigm that protects individual autonomy and privacy in the process of data collection and utilization and respects the human context contained in data.

Data imperialism and Apple

Apple uses 'privacy' as a core marketing element. Screenshot of Apple homepage

The metaphor of data as oil or natural resources has driven the massive flow of data collection. On the opposite side of this flow stands one company: Apple.

Apple has shown a differentiated approach from other big tech companies based on the philosophy that “privacy is a human right and a civil right.” In 2021, Apple announced the “App Tracking Transparency” policy. It requires all apps to obtain explicit consent before tracking user data. This sparked strong opposition from advertising-based companies like Facebook and Google.

Apple’s AI system, “Apple Intelligence,” also emphasizes personal information protection unlike other AI services. Apple introduces that “the right to privacy is a fundamental human right everyone should enjoy and one of our core values.”

In fact, Apple Intelligence runs entirely on-device without connecting to the internet. It can complete tasks without data leaving the device. Only for complex computations that cannot be handled internally does it connect to external (private cloud computing) to process user requests.

Of course, personal information protection is key even in this process. Apple emphasizes that “transmitted data is used only for the user’s request, is never stored, and no one can access it.”

In a situation where some claim “Apple is a latecomer in the AI competition,” Apple’s strategy might seem like a “mistake” at first glance. Privacy protection conflicts with data collection and AI development. Focusing on privacy could hinder data collection and slow AI advancement.

Nevertheless, Apple has made privacy protection a marketing point. Paradoxically, this shows that privacy protection can also be a business. Although it may lose advertising revenue in the short term, it can gain advantages such as brand value enhancement and securing loyal customers in the long term. With privacy-related regulations increasingly tightening, Apple’s proactive response could be a strategy to gain unique competitiveness in the future market.

Data is important, but personal information is also important. There is no need to be trapped solely in the claim that “data is the new oil.” Along with this, we must value our personal information more highly ourselves. Our personal information is not a passive record to be “extracted.” It is private records, property, and value of each individual. Providing my step count data and receiving a car insurance discount was by no means free. What Apple CEO Tim Cook said in a 2018 media interview remains valid today.

“When an online service is free, you’re not the customer. You’re the product.”

Next Series Preview (Every Saturday)

(20) The thought that K-AI is ‘Korean nationalism’ (03.01)
(21) AI suffering from genetic diseases (03.08)