The all-seeing spy in your pocket

What does your smartphone know about you and who does it tell? Adriana Hamacher delves into the data goldmine in your pocket and raises some concerns

Always on, and monitoring a user’s every move, swipe, tap and pause, our smartphones hold a data narrative of immense worth to big tech and other industries. Data that users knowingly share, such as age, gender and interests, is just the tip of the iceberg, and mostly used to fuel targeted ads. But beneath the surface lies a far more intricate data trove that’s attracting interest from a range of industries seeking to capitalise on its insights.

Apple Intelligence, a new AI-powered system marketed as a tool for hyper-personalisation, is now sparking debate over its ability to track app activity and push the boundaries of smartphone surveillance. While its launch highlights growing concerns over privacy, it also underscores the expanding role of smartphone data in shaping industries such as finance, where its analytical power is both transformative and contentious.

Data collectors

Smartphones have long been central to the economics of data-driven advertising – collecting and sharing insights that power both personalisation and profit.

Most data is anonymised and acquired from various sources. For example, many mobile applications collect users’ location data, which is then aggregated and sold to data firms to offer insights into consumer behaviour, foot traffic patterns and other metrics valuable to retailers and investors. Users consent to this data collection through app permissions, although they may not be fully aware of its extent.

From tracking spending habits and sleep patterns to logging driving speeds and accessing contact lists, the apps we use can quietly gather vast amounts of data. Sophisticated identifiers underpin this system, linking disparate data points to create detailed profiles.

Behavioural insights, such as swiping patterns, typing speed and screen pressure, are increasingly fed into AI systems, adding even greater depth to profiles.

A recent study in Nature Digital Medicine demonstrated that smartphones, through passively collected behavioural data, can predict mood changes weeks in advance. For example, reduced activity on social platforms was linked to declining mental health, highlighting the potential of everyday digital habits to signal underlying issues.

But, considering that every tap, swipe and pause can reveal habits, preferences and even personality traits, it raises important questions about who controls this data, how it’s used and what happens next.

It doesn’t just happen

Regulations governing personal data differ sharply between Europe and the US. In the UK, data is either ‘special category’, which includes sensitive details such as ethnicity, religion, health and biometrics, or non-sensitive, which companies can handle more freely. While transaction data isn’t inherently sensitive, it can reveal protected characteristics when enhanced with new details.

From tracking spending habits and sleeping patterns to logging driving speeds, the apps we use can quietly gather vast amounts of data

In June 2024, The New York Times exposed how seemingly harmless smartphone apps collect and sell driving data to insurers. Apps such as Life360, MyRadar and GasBuddy – promoted as tools for family tracking, weather updates and fuel savings – analyse behaviour such as speeding and distracted driving to generate ‘driving scores’. These scores are then sold to insurance companies to influence premiums, often without users fully realising the implications of their consent.

There are particular dangers for data-driven marketing in finance, as Revolut has discovered. In February 2019, the bank’s advertising on London Underground asked: “To the 12,750 people who ordered a single takeaway on Valentine’s Day. You OK, hun?” Widely criticised for ‘single-shaming’, it also sparked outrage over its casual use of transaction data.

However, investment managers are increasingly turning to alternative data (alt-data), which includes credit card transactions, social media commentary, product reviews and satellite imagery. A 2023 survey by Lowenstein Sandler, a US-based law firm, found that 62% of investment professionals now use alt-data, up from 31% the previous year.

Users reported spending between $1m and $5m per year on alt-data in 2022, and revenue generated by providers such as Eagle Alpha is projected to overtake that of traditional data services such as Moody’s by 2029, according to Deloitte.

“Whether it’s in lending, or credit decisions, whether it’s in KYC (know-your-customer), CDD (customer due diligence) or transaction monitoring or fraud, [there are] many, many different options and opportunities for us to pursue,” said Marnix van Stiphout, Chief Operations Officer at Dutch bank ING, speaking at Money 20/20 Europe in October last year. “The real question is, can you focus on the ones that really make most sense? Because there’s no shortage in terms of ideas. There is, however, a shortage in terms of capability [and] quality of people.”

Donald MacKenzie, a sociologist at the University of Edinburgh known for his analysis of the interplay between technology and finance, agreed on the potential. “I’m sure the data is potentially very useful,” he said. But he added that “banks are largely out of the loop. Although they can, of course, buy data from the brokers”.

AI is expected to deepen banks’ reliance on major US tech firms, with many already dependent on their cloud services to process sensitive data. Executives at Money 20/20, including Joanne Hannaford, Head of Technology Strategy at Deutsche Bank’s corporate bank, see this relationship as unlikely to change. “AI requires huge amounts of compute [computing power], and really the only way to access that compute sensibly is from big tech,” she explains.

Hannaford also emphasises that big tech companies are more likely to become partners than competitors to banks. She points to significant barriers to entry – such as cost, time and complexity – that make partnerships a more practical path forward.

Scaling the walls

Despite safeguards, widespread consumer concerns about data collection practices by major tech firms such as Google persist. While Google’s reliance on advertising revenue makes its need to carry out data-mining clear, Apple has built its brand on privacy. But its evolving approach to data, particularly with tools such as Apple Intelligence, reveals a more complex relationship.

Big tech companies are more likely to become partners than competitors to banks because of the significant barriers to entry

Launched at the end of 2024 in the US on the iPhone 15 and 16, Apple Intelligence is the first Apple-branded GenAI product in the market. The company has partnered with OpenAI to integrate ChatGPT into the iOS operating system, a significant milestone for the company because its introduction will have widespread effects on every iPhone feature. Notably, it allows the technology to access and analyse app data, including sensitive banking, financial and location information.

Consumer insights analyst Giovanni Siracusa warns that Apple Intelligence, which launches in Europe this year, could fundamentally alter the company’s privacy model. “Everything that might be used internally in a cell phone will be integrated with ChatGPT and OpenAI to produce so-called Open Intelligence, and this is a risk,” he says.

The line between financial services and advertising giants is rapidly blurring as data becomes the currency of competition

He highlights the dangers of lowering barriers between apps, allowing third-party manipulation of aggregated sensitive data. This could lead to entry barriers for services such as Buy-Now-Pay-Later based on user habits, creating inequities in access. In a largely unregulated field, Siracusa warns, the potential misuse of data raises serious privacy concerns.

Apple claims that its AI processes data in its ‘Private Cloud Compute’ system to deliver personalised experiences while ensuring user privacy. The company insists that it does not use, view or store personal data on its servers.

Silver linings

Despite growing scrutiny of data practices, the market for big data is forecast to more than double in size within the next decade, according to a November 2024 report by Allied Market Research.

To address rising demands for greater transparency, companies are exploring ways to anonymise data without sacrificing functionality. Google, for example, is working on de-individualising mechanisms, in the form of the Android Privacy Sandbox, expected to launch this year.

Beyond these challenges, the societal potential of smartphone data is immense. It can help track the progression of diseases such as cancer, improving patient outcomes, or monitor infrastructure health by using smartphones in vehicles to detect bridge integrity issues, alerting authorities to maintenance needs.

And, as the battle for data-driven innovation heats up, some fintechs are taking on the ad-driven tech giants. Revolut, undeterred by past controversies, plans to sell advertising space on its app, aiming to become a media and advertising platform.

Similarly, Zilch, a UK-based fintech backed by eBay and Goldman Sachs, uses targeted ads from transaction data to subsidise zero-interest loans. “We’re actually an ad platform that’s built a credit proposition on top of it,” Chief Executive Philip Belamant told the Financial Times last year. The line between financial services and ad giants is rapidly blurring as data becomes the currency of competition.

Adriana Hamacher is an independent researcher in human-robot interaction and an award-winning writer specialising in emerging technologies. Her work has been featured by the BBC, Wired, Mashable and other media outlets