Foreword

Digital technology is rapidly changing the world. The number of mobile banking users is forecast to double between 2015 and 2019 to 1.8 billion, representing more than one quarter of the world's population. As one of the world's largest banking and financial services organisations, serving more than 37 million customers across 70 markets, we need to anticipate and be ready to respond to that change.

Our Trust in Technology research explores public opinion about the new technology banks and consumer facing businesses are developing, and its impact on daily lives. In this international study we examine people's awareness and understanding of new technologies; their trust in them and the impact of this on adoption rates; and what people think the future looks like in terms of new innovation. The report explores some of the top technology topics of the moment, like artificial intelligence, biometrics, digital wallets and data driven nudge applications.

We will apply the insights this research provides to our own business, as well as sharing them broadly with the financial services industry and other stakeholders.

Digital demands vary from customer to customer and market to market, but the long-term direction is clear: customers should be able to bank with us when they want, in the way that they want. At HSBC, we want to apply proven digital technology to make our customers' experiences simpler, better, faster and more secure. We will continue to adapt as their needs change, to provide banking services on their terms.

John Flint

John Flint

Global CEO, HSBC Retail Banking and Wealth Management

Why we need to study the nature of trust

Technology is advancing at a dazzling rate. Life expectancy is hitting record highs thanks to ground-breaking new processes. A team at Harvard Medical School has grown heart tissue from adult skin cells. The prospect of transplant organs grown from scratch is on the horizon.

The success of Amazon Echo and Google Home smart speakers shows there is an appetite for Artificial Intelligence (AI) assistance in the home. Autonomous cars are being test driven. Sometime in the next decade it will be normal to be chauffeured by an AI.

Financial services are being re-invented by new technology. Apps make it quick to send money and check accounts. Online consultation services nicknamed robo-advisors are bringing expert guidance on investment and pensions to a wider audience.

With each new piece of technology, the same fundamental question arises: can it be trusted? Price and availability matter, but nothing is adopted unless the target user feels able to trust it.

But what are the factors which determine what we trust, and what we reject?

With each new piece of technology, the same fundamental question arises: can it be trusted?

This report is an exploration of the state of trust in technology

As industry and society is transformed by technology, HSBC decided to explore how people feel about the changes, in particular how the concept of trust might help or hinder this evolution. The aim was to try to identify the barriers to progress, the appetite for change, and the benefits that greater trust in technology might unleash.

To understand those questions we commissioned research among over 12,000 people in 11 countries, interviewed scores of experts, and reviewed the results. This shortened report (available in full at hsbc.com/digital-and-innovation) reveals the level of trust in biometric security, attitudes to radical new financial services, and what we really feel about entrusting our lives to a robotic intelligence.

It has become clear that the more technology enters our lives, the more critical the issue of trust becomes. A FinTech start-up can't win customers without winning their trust. A bank can't keep customers without being trusted. In order to deliver life-enhancing new products they need to know what trust is, and how to win it. This report is offers some suggestions on how trust in technology can be achieved.

It has become clear that the more technology enters our lives, the more critical the issue of trust becomes.

The state of trust in technology today

Technology continues to develop at a rapid pace with faster, simpler and more secure innovations being launched on an almost daily basis. But what is the public appetite for creative new tech?

Our research shows a lack of trust and understanding is holding back the adoption of new technologies.

By default, there is a high level of social trust. People give each other the benefit of the doubt, with 68% saying they will trust a person until proved otherwise, with more than twice as many believing the majority of people are trustworthy than not. The outlook on life is positive, with 65% confirming they are always optimistic about their future, and half saying "things that are bad normally turn out okay in the end", triple those disagreeing.

This warmth extends to using technology, with 76% saying they feel comfortable using new technology, and 80% saying they believe technology makes their life easier.

But when respondents were asked about fields of emerging technology. The sunny outlook quickly fades.

Artificial intelligence is a booming field, with engines such as IBM Watson and Wipro Holmes able to diagnose cancer, analyse retail data, and communicate through ordinary spoken and written language. However, AI struggles to win support. Only 8% would trust a machine to offer mortgage advice, compared to 41% trusting a mortgage broker. For context, this is 2% lower than flipping a coin for financial advice and 1% lower than using a horoscope.

Robotics is advancing through links to AI. How about allowing a robot to conduct a surgical operation? Only one in seven would trust a humanoid robot programmed by experts to conduct the operation on them, just five percentage points ahead of the rather unnerving prospect of a family member or friend doing the operation under guidance.

Biometric security is another field with huge potential. It's possible to use fingerprint, voice, iris, and facial patterns to confirm an identity. Security experts regard biometric as a potential breakthrough in the fight against identity theft.

However, trust is yet to be achieved in biometric technologies. Iris recognition is trusted by 26% to replace alphanumeric passwords, a combination of letters, symbols and numbers, with voice recognition and facial recognition trusted by 18%.

Winning and breaking trust

The results reveal a fundamental tenet of trust in technology. Unlike people – who we trust until given cause to think otherwise – trust in technology is earned, not given. Caution is the default.

In depth interviews conducted as part of this study looked deeper at the thinking behind trust. Consumers were interviewed across international markets, and the results were then discussed by a panel of technology experts.

Trust in people produces testimony brimming with emotion and enthusiasm. An interviewee from Hong Kong produced a photograph and said, "My best friends since secondary school. I truly trust every one of them due to our deep bonds. Though we're in different universities and experiencing different university life, we are also the best listeners to each other and truly consider each other as brothers."

By contrast, trust in machines was expressed as transactional. Performance is what matters. A Chinese interviewee expressed their logic on what to trust: "It [Trust] is being sure that the thing will not fail or disappoint you with its quality, operation and/or handling. Being sure that it will do what you expect it to."

The principle that trust in technology is built over time, based on hard evidence, explains the hesitancy around new products and services. Chatbots, AI and biometrics are developing fast, and are yet to prove functional reliability to doubters.

Overall, many consumers are still unaware that emerging technologies even exist.

More familiar concepts generate higher levels of trust. This is proven by perceptions of fingerprint technology. Fingerprint identification has more than a century of use by police forces. It is a familiar method of identification. Fingerprint recognition is already used by 21% of respondents internationally, rising to 40% of Chinese people and 31% of Indians. The status of the technology reflects this familiarity, with 46% globally saying they trust fingerprint recognition to replace passwords in the future. Financial services companies will need to cater to this demand: 38% said they regard fingerprint login as an important or essential service.

Current attitudes to emerging technologies should therefore be taken with caution. Trust can be won as each technology becomes part of normal life and develops a track record.

"This is a growing issue," says Niall Cameron, Global Head of Corporate and Institutional Digital at HSBC. "You entrust your data to one service provider who then passes it onto another without your knowledge. You could be wearing a fitness tracker and input your diet, whilst it monitors the exercise you do. This is great for improving your fitness. But what if the data is passed to an insurer who looks at what you eat and drink, and thinks you are not active enough hikes your premiums, would you be happy with that?"

Chart 01

Figure 1: I have never heard of this technology

The optimism of the East

Attitudes to technology vary from nation to nation, and the survey reveals the key factors. Openness is correlated to social and economic change, with rapidly developing markets in the East far more experimental with innovation than in the leading mature economies of the West.

China and India are the most open to new technologies. France and Germany are the most cautious of the 11 nations surveyed. For example, 41% of Chinese and Indian respondents said they were "likely or very likely" to trust a hologram assistant to help them make choices around investments. In France and Germany the figure is 7%.

Figure 2: I believe technology will make the world a better place

Chart 02

The research sheds light on why trust in Western societies is lower than in the East. A Canadian interviewee explained their reticence to trust corporations: "It's hard to trust blindly these days, especially with technology and companies. You hear about trust being violated all the time-personal information being compromised, shared, hacked, and breached. Giving trust to companies is a big leap these days."

An interviewee from Mexico explored one of the causes – repeated breaches of trust: "I've had my trust broken many times when buying products and services that promise something specific, and then when you have them in your hands and operate them, it is all just marketing."

The vocabulary of participants illustrates the mindset in Western countries. There has been an erosion of trust combined with an empowerment in technology. The result is greater self-reliance.

Chart 03

Figure 3: Who do you trust to keep you safe?

In markets where there is greater trust in institutions, there is a greater hunger to experiment, leading to higher familiarity and trust.

In India consumers send payments via WhatsApp. In China, new payment services such as Alipay and platforms such as WeChat have changed the market. Kevin Martin, Head of RBWM, HSBC Asia-Pacific, says he has seen this first hand, "China's consumers are years ahead of their counterparts in many developed economies in terms of how they ship and pay for what they buy."

How financial services organisations should react

The principle of withholding trust in technology until confidence is won is highlighted in the finance sector. Caution is amplified when money is at stake. The priorities are stability and rigour. The attractions of new FinTech services rate below these mandatory factors.

Just 16% of respondents said they would switch bank for one with better technology, although the number leaps in markets such as the UAE 29% and India 32%.

Old fashioned complaints are far more likely to trigger a switch of bank. Being hacked into and having money stolen (52%), and finding out the bank does not have the customer's best interests at heart (36%), are more popular reasons to switch.

Consumer priorities are basic. Banks need to keep money safe, with 87% calling this important or essential, and the same number demand security of personal data.

Chart 04

Figure 4: My bank is "good enough for what I need it for"

Banks are conservative adopters of technology, and consumers appreciate that prudence. There's a view that banks are not the right organisations to experiment – 61% agree that banks should only use technology which has been tried and tested. Regulation plays a role in guaranteeing consumer faith in banks, with three-quarters saying it is essential or very important that new technology is independently regulated for security.

The survey explores how consumers build trust in technology. At first trust is withheld, and builds over time based on evidence. Trusted brands and institutions can accelerate that transition.

The high degree of trust in banks may provide a strong foundation for them to promote new financial services, provided they respect the level of rigour and caution expected by consumers.

"You hear about trust being violated all the time-personal information being compromised, shared, hacked, and breached. Giving trust to companies is a big leap these days."

Interviewee in Mexico

Love, lust, and convenience

A common assumption in the theory of trust is that you either have it or you don't. This reductionist approach ignores the different intensities of trust. Love for humans tends to be deep. Emotion pulls us together, and when trust breaks the consequences are painful.

Human trust can trigger heart-warming and colourful responses. One of our interviewees from Singapore was asked who they trusted the most, and answered: "Trust is a precious relationship matter which ties two people into the same common circle and walk into same direction. To me, trusting someone means you feel secure to share your true feeling, your own self at any time. You are free from any worries or embarrassment in front of the trusted person."

Trust in machines is functional. It lasts so long as the device or service delivers a result. An interviewee in Hong Kong said that, "To trust something means to believe something that is true or correct or reliable that you can rely on it. Like my mobile phone or the lock of my house."

"Trust is a precious relationship matter which ties two people into the same common circle and walk into same direction. To me, trusting someone means you feel secure to share your true feeling, your own self at any time."

Interviewee in Singapore

Figure 5: The different types of trust can be mapped against time and depth

Chart 05

Where trust is deep and long-term there is love. Family members and close friends are in this quadrant (below). Long-term institutions for whom we have modest levels of trust are necessary features in our lives – there's a quadrant to reflect those values. Some services we use also to achieve a functional goal, but we take them to heart a little more. Search engines and online shops are functional in this way – we rely on them, and trust them to work, but we might abandon them if something better came along.

Then there's the zone of frivolity – lust. Start-ups with unproven business models might catch our eye for a moment. Lust is shallow and short term. Buying a new gadget can feel like a quick fling – exciting, and experimental.

Chart 06

The goal for any innovative brand is to move out of the lust zone by building deeper levels of trust. Long term use by customers can achieve this. A Canadian interviewee expressed their personal experience: "I have been using PayPal for a very long time, and I really enjoy the convenience and constant support from the company. Facebook I trust because of the number of times I have used it."

Chart 07

Here's the quadrant for tech companies.

Currently only humans sit in the love zone. It's possible to move a brand closer. The key is to foster the same "human" elements such as compassion, understanding, integrity, and safety.

Financial AI that delivers some artificial humanity will be able to deliver genuine benefits to consumers, deepen relationships, and build trust over the long term.

Cure security and privacy worries and adoption will increase

Building trust takes time. It requires familiarity with a product, and repeated results. As the understanding grows, trust builds.

The process can be fragile in the early days. The research examined why so many technologies are struggling to get through an initial period. Respondents were asked for their biggest worries in life. The top-ranking concerns were "personal data being leaked" followed by "bank account hacking" and "debit or credit card cloning". These outrank a fear of serious illness or being burgled. In finance, security of assets, and security of personal data, are both rated by 87% as essential or important. It's clear that security and privacy are critical issues preventing a higher adoption of devices.

How to address security

In theory, there's no need for consumers to be anxious about security. Tools to protect ordinary activities are comprehensive. Sophos anti-virus principal research scientist Chester Wisniewski says, "Practice good cyber-hygiene and you shouldn't get too worked up. Business as usual."

The problem is that consumers often aren't always taking the right precautions. In the survey, only 33% said they protect their devices with up to date security software. Only one in three uses different passwords for different banking products – in breach of guidelines.

Consumers often worry about the wrong things. Biometric security is regarded by industry experts as superior to alphanumeric passwords. But adoption is being held back by fears that that the central database could be hacked. This a misunderstanding of the authentication process. A system called "tokenisation" means the fingerprint scan stays on the phone. There's no database to hack.

Multi-factor security can address concerns. Multi-factor means two or more security systems are used together. For example, a text message can be sent to the user's phone with a one-time entry code to confirm possession of the device. The Ipsos MORI survey shows 46% of Chinese respondents use third-party authentication.

The ideal multi-factor system takes the initiative away from the users. This is the goal of behavioural analytics. Keyboard typing rhythms, voice pitch, mouse movement, walking gait, and speed of swiping can be tracked to create a "digital fingerprint". Swedish company BehavioSec looks at the angle a user holds a phone, how much of a finger touches the screen, and how quickly fingers move when typing.

Darryl West, Chief Information Officer, HSBC, says biometric and behavioural analytics offers a step change in convenience and security: "My objective when I finish my career in this industry is to eliminate passwords completely."

Data privacy is a major concern

A further worry for consumers is fear that data is being shared. The poll shows data privacy is rated as important as protecting finances.

A survey of American and British adults by Rackspace Hosting found 51% cited privacy concerns as a barrier to adoption of wearable technology. If these privacy issues are addressed the gains could be huge.

The way forward

Add up these complications and it becomes clear why consumers worry about security and privacy.

Fortunately, the right technologies exist to deal with the main worries.

With the right information and tools, fears over negative outcomes can be reduced, giving consumers the opportunity to enjoy the benefits of new technologies.

"Practice good cyber-hygiene and you shouldn't get too worked up. Business as usual."

Sophos research scientist Chester Wisniewski

The rise of Robo-Advisors

Robo-advisors are one of the most exciting new technologies in industries including financial services. These consultants take the form of a web service. The robo-advisor poses questions to the user, and then recommends the best course of action based on the evidence.

Popular fields include pensions, investment and savings advice.

The survey set out to establish the level of trust in robo-advisors at this early point in their development.

Overall 19% would trust a robo-advisor to help make choices around investments, rising to 38% in India and 44% in China. There are sceptics. Just over half of people would be unlikely or very unlikely to trust a robo-advisor, with trust falling to 9% in France and 6% in Germany. Giving the robo-advisor a user-friendly holographic make-over offers no gain, the trust levels are identical.

There are pros and cons of the robo-advisors. In order to increase trust in robo-advisors it will be necessary to communicate the advantages they bring. Here are the seven main arguments in favour of these digital consultants.

1. Harness the wisdom of multiple experts

Robo-advice can be programmed by not just one expert, but a committee of experts each contributing their specialist knowledge. The advice is inspected, reviewed, and tested to ensure the users receive the highest possible quality of advice.

2. Low cost

The cost of robo-advice often undercuts human consultancy fees. This means savings for users. It also makes financial advice available to a cohort of young and lower-income consumers who were previously priced out of the market.

3. Friendly user experience

Robo-advisors are convenient for people who live in remote areas, or work unsocial hours. The online format is an appealing alternative for anyone too shy to book an appointment with a human advisor.

4. Frequent upgrades

For robo-advisors constant improvement is a way of life. The programmers can add new functions at will.

5. Real time information

A robo-advisor can be plugged into market data to match advice to market changes such as changing mortgage deals. A robo-advisor is purpose built to examine the full range of market options, and adapt advice in real time.

6. Rebalancing

Robo-advisors are formidable in investment scenarios. They can track portfolio values, identify potential changes, and implement whatever strategy they are programmed to follow.

7. Works together with humans

Robo-advisors take the simpler and more onerous duties, such as offering early consultation to would-be investors, and answering basic questions. This liberates human counterparts to focus on specialist activities.

"In order to increase trust in robo-advisors it will be necessary to communicate the advantages they bring."

Charlie Nunn, HSBC Global Head of Wealth Management

"There is an advice gap in financial services. Unless you have a certain amount to invest asking a professional adviser doesn't make sense, and that is where a robo-advisor can really help a big chunk of society. The technology is actually quite simple, using 'if then' decision logic. The results are crafted by our best advisors and consistent with the regulatory framework, so everyone gets a very high standard of advice."

Charlie Nunn, HSBC Global Head of Wealth Management

Question time: What do you trust?

The next generation of software and hardware will be more powerful and creative than anything we've seen to date. It will have the ability to supplement human endeavours in bold new ways, and even challenge human performance in a growing number of fields.

The arrival of more powerful systems will provoke profound questions. Here are two of the most challenging conundrums. In both cases, human users need to assess how far they trust technology to complete a task.

Question 1 - Would you trust a machine to set you up on a date over a family member?

Can an algorithm beat human judgement in selecting a romantic partner? The research says not. A friend is trusted by half of respondents to find the right person, and a third would trust a family member. Only 8% trust a robot programmed by experts in relationships.

Could we get respondents to think again? In fact, we can do a deep analysis of the machine's chances. Christian Rudder founded dating site OK Cupid, now owned by IAC, which also runs Match.com and Plenty of Fish. He crunched the numbers across 55 million users and published the results. Looks, he discovered by looking at satisfaction on blind dates, are irrelevant.

The same is true in the partner checklist. Priorities such as religion, politics, and smoking are usually rated "mandatory". Yet two oblique questions have superior predictive power: Do you like scary movies? and Have you ever travelled alone to another country? Three-quarters of long-term couples brought together answered the same way. Rudder notes, "People tend to overemphasise the big, splashy things…[but] sometimes they don't matter at all."

The message stretches beyond dating. Machines can use the power of big data to draw insights into our desires that even we are blind to.

Question 2 - How much should we use nudge technology to improve behaviour?

Economists Dick Thaler and Cass Sunstein invented nudge theory, and the pair are brimming with examples of how effective it can be. In their book 'Nudge' they describe a pension scheme which harnesses the concept of "loss aversion" - the idea that consumers hate losing something more than they enjoy the gains. Loss aversion explains why pension contributions are lower than they should be. The pair created a pension plan in which no contributions are made until a pay rise. Then a percentage is paid in. The saver therefore never sees a reduction in income. The idea was a triumph. Contributions rose 200%.

When used effectively a nudge can produce net benefits. An app called SmartSave developed with a FinTech company called Pariti helps increase the amount they save. The app can be linked to any account, and automatically diverts money into a savings account according to user-defined rules. For example, purchases resulting in a few pennies over or under a round pound results in the change going to the savings account. No action required. It prompts users when it's "safe to save". As a result, money accumulates in the savings account without effort.

Does that mean there's an open license for nudging? The author of the theory isn't too sure. Cass Sunstein published a follow up work called The Ethics of Influence looking at the dangers of over-use. He raises the issue of consent. There the question of responsiveness - stress can make people more reactive to nudging. Might it lead to lower-income workers getting harder nudges than the well off? And the issue of incentives – who is really benefiting?

Sunstein states, "Certainly choice architects should be focused on the welfare of choosers, rather than their own". He notes the extreme sensitivity to issues of manipulation, particularly in nations with a history of poor human rights.

Nudge theory is a potent tool. But if used to excess then the good it can do may be curtailed.

Machines can use the power of big data to draw insights into our desires that even we are blind to.

The path to high trust

Trust is a critical concept in the adoption of new technology. High trust means early adopters are prepared to play with new concepts. Breakthrough ideas get trialled and validated. In financial services, high trust means brands can roll-out new services which improve lives and open new opportunities. It's a virtuous circle. High trust means consumers adopt better security tools like biometric identification, leading to fewer breaches.

This report shows there is a need to focus on improving trust, and offers insights into what can be done. Here are the five essential lessons for policy makers and brands.

1. Invest in education

There is always a gap between technology and public awareness, and education is the key to minimising this gap. Great work is currently being done. Google India is working on educating the public on safe internet browsing. The FIDO Alliance is promoting understanding of tokenisation in the biometric industry – essential if consumers are to understand there's no central database to be hacked. Even small gestures can help. The research showed trust in biometrics rises 6 percentage points after a short briefing. The march of technology is remorseless. Education will ensure no one is left behind.

2. Kitemarking

Kitemarks are a simple way to display compliance with official standards. Progress is already being made by organisations such as the GSMA, and the Internet of Things Security Foundation. The goal should be to provide an intuitive set of symbols to guarantee minimum levels of security and performance.

3. Simplified interfaces

User behaviour is to a large part dictated by the structure of the user interface. If good choices are easy to find and select, and poor choices are excluded or hard to find, then it is possible to increase performance and trust. Interfaces should be standardised whenever possible.

4. Accelerate biometric and behavioural security

Biometric security offers a clear way to speed up authentication whilst reducing inconvenience. Behavioural analytics is a complimentary technology, typically working without the consumer needing to do anything at all following enrolment. Multi-factor security should be promoted.

5. Regulation, including transparent algorithms

The research shows emphatic support for independent regulation (75% in favour). This tallies with other sectoral surveys, such as an IOActive poll of IT professions showing 83% support. It may be necessary to establish an independent watchdog to police AI, and offer redress for people who believe they have been discriminated against in public contexts.

The march of technology is remorseless. Education will ensure no one is left behind.

These measures directly address the trust deficit. The impact could be significant. The world is at risk of creating distinct classes of technology users, with less-able users increasingly left behind.

A few thoughtful steps, like promoting biometric security and developing simplified interfaces, can spread the benefits of technology to the widest possible number of people. Education can help even novices get involved.

This is more than economic prudence. The human race is moving into an era where technology shapes our lives. It's a journey we should be taking together, so we can create a world in which we all benefit from the fruits of innovation.