Welcome back! If you’re new, sign up here to receive this free email every week.
Facial recognition technology is popping up in the news everywhere. In London, people are in a tizzy after the Financial Times (paywall) reported that a property developer is using it to track thousands of people in a busy part of the city. Another report claimed a massive database of biometric data was left relatively out in the open on the internet.
These stories capture two of the worries people have about biometric data, and facial recognition in particular: that companies or governments could use this information to surveil us, and that hackers could steal it. Unlike a password, you can’t change your face after the digital template of your biometric markers is stolen. It still feels creepy to be recognized by a machine, and it’s frightening to think of systems logging your presence without permission or acknowledgment.
These risks are real, but are also easy to overhype and sensationalize. In some instances, for example, facial recognition isn’t creating a substantial new database of information. It’s just automating and making use of photographic data that was being stored already.
Take airport security (paywall), where facial recognition is being used to speed the boarding process for international flights from the US. Instead of an airline employee comparing a passport to passenger’s face, a computer uses an existing database of photos, such as passport pictures. Customs and Border Protection told the Wall Street Journal that it doesn’t allow airlines to to store the data. The process appears faster and less error-prone than the manual method.
Something similar is in play with Alipay’s face payment systems. Last month, a story suggested that face payments could give QR codes a run for their money in China, and my colleagues have written about Ant Financial’s Alipay adding beauty filters to its face-pay service. This can be faster for merchants, and helpful for elderly people who might not be comfortable with payment apps.
Financial institutions already have to know who their customers are as part of money laundering regulations, meaning banks may hold photo identification of customers in order to comply. (When I signed up for my last bank account, it used a picture of me holding my passport.) In some cases, it may be more accurate to say that facial recognition is drawing attention to information that was already being stored, rather than building up collections of new data.
Another big worry is that biometric data can’t be changed if it gets stolen. This information will become more valuable to criminals over time as we rely on it more, according to Darktrace director of technology Andrew Tsonchev. It could be used for high-value targets, such as impersonating bank employees.
But right now we typically rely on passwords, which are obviously lousy, and most of them have been stolen anyway. If hackers have a database full of facial recognition templates, would they have to print 3-D models of their victims’ heads to be able to break into their accounts? These are serious risks, but there seems to be much less discussion about how these vulnerabilities could actually get exploited.
Apple’s iPhone X demonstrates some of the steps that can keep the information secure. The user’s data is kept only on the phone (it isn’t uploaded to the Cupertino’s company’s servers), and the biometric information is encrypted and contained in firewalled hardware within the handset.
Michael Veale, a researcher at the Alan Turing Institute, pointed out to me that while facial recognition is getting more attention now, it has been around for decades. Rights and regulations are still evolving and being debated, but places like the UK and the rest of the EU have laws in place that put some guardrails around how this data is used by corporations, police, and intelligence services (in Europe, the rules are different for each of these entities).
There’s no question that facial recognition is vulnerable to abuse, as an investigation by Georgetown Law into police use of this technology found. Making sure that people can truly opt out of having their data captured is a key consideration for maintaining privacy and freedoms. (Just ask Hong Kong protestors.) The trick, of course, is sound regulation. San Francisco has banned facial recognition technology, but Veale notes that tech is already being developed that can figure out a person’s identity based on their walking gait or heat signature. While the California regulation received a lot of attention, it’s not clear it will provide the intended result for the public. As machines become smarter, regulation will have to get wiser, too.
This week’s top stories
1️⃣ PayPal’s chief technology officer told the Financial Times (paywall) that “India is critically important.” The company is increasing employee headcount and looking for acquisitions there.
2️⃣ Tech companies in China are wooing elderly users, according to The Economist (paywall). A survey found that only 17% of seniors frequently pay for things using mobile phones, and nearly half had never done so.
3️⃣ China says it is close to starting a digital currency. It’s still unclear how, and whether, it would resemble cryptocurrencies that change hands outside the traditional banking system.
4️⃣ HSBC is making a push in the US personal loan market. The bank is using fintech Amount‘s lending platform.
5️⃣ Monzo is offering short-term loans. Part of the promise of neobanks is to bring tech company-inspired business models to finance. But this sounds a lot like the old business model (paywall).
The future of finance at Quartz
More African countries are looking to tax mobile money transactions. Governments there are looking beyond mining companies and oil multinationals for tax revenue.
Fintechs are making it cheaper than ever to send money across borders. But our research on PayPal’s Xoom service shows you still have to do your homework.
JPMorgan Chase forgave its Canadian customers’ credit card debt. As the bank closes its Canadian operations, it appeared to be cheaper to write off the credit than sell the assets or continue collecting on the debt.
Always be closing
- Curve is launching a seven-figure crowdfunding campaign. The fintech’s customers will be able to invest as little as £10 ($12).
- Fintech investment in the UK and Germany about doubled during the first six months of the year, much of which was driven by neobanks. Chinese dealmaking stumbled, while US investment increased 60%.
- Bond-trading platform MarketAxess is buying LiquidityEdge, a US Treasuries marketplace, for $150 million.
- Figure Technologies is raising more than $100 million at a $1 billion valuation. The company was founded by Mike Cagney, the former CEO of Social Finance.
- Atom Bank, Iwoca, Currencycloud, and Modulr Finance were awarded £10 million each by the Banking Competition Remedies fund, which was set up after Royal Bank of Scotland’s bailout.
- UK-based Flatfair, which allows tenants to rent a new property without a deposit, raised $11 million in a round led by Index Ventures.
I hope your weekend is invigorating and profitable (pick your own metric). Please send facial-recognition templates, tips, and suggestions to firstname.lastname@example.org.