There’s a paradox at the heart of this moment in the information age: We’ve never had access to more data, yet we’ve never trusted data less.

More than a decade into the big data era, information has become a potent currency—the “principal driver of economic growth and wealth in this era,” as Bob Fay, the former Managing Director of Digital Economy Research at the Centre for International Governance Innovation, asserted in 2023. As organizations work to capitalize on the increasingly dazzling potential of machine learning and artificial intelligence, a plentiful supply of good data can be as valuable—if not more—than any physical asset. It is now in any company’s best competitive interest to be able to gather and analyze as much intel about its customers as possible, and the tools of the digital age—from smartphone apps to online databases—make it relatively easy for them to do so.

However, this profligation of information is doing little to put the public at ease. People are worried: According to a 2023 survey from the Office of the Privacy Commissioner of Canada, 93 per cent of Canadians are concerned about the protection of their privacy, and fewer than half feel equipped to understand how new technologies affect their personal information. They’re also wary: The 2024 Edelman Trust Barometer showcased a pervasive societal skepticism in the ability of institutions (including businesses) to securely handle sensitive information, while the 2024 Telus Data & Trust Survey found that nearly three-quarters of Americans and Canadians are now concerned with how organizations handle their personal details. People now care nearly as much about a corporation’s ethics and trustworthiness as they do service and price. When something like a data breach occurs, they are quick to bristle and slow to forgive.

In this context, companies pursuing the spoils of data-driven innovation cannot assume the trust of their customers and other stakeholders—they must actively earn it. “We know that our customers are sharing with us a massive amount of data, and that their daily lives are intertwined with our digital ecosystem,” explains Pam Snively, Telus’s chief data and trust officer. “We take that relationship very seriously.”

In recent years, Telus has implemented a series of enterprise-wide education initiatives, processes and policy measures meant to enhance the trustworthiness of its offerings without compromising its ability to innovate. “We’ve really made an effort to be more transparent in how we’re using data,” Snively says. “And we’ve done a lot of work to apply rigor and consistency in all of our data activities so that they are sustainable and worthy of customer trust.” 

What, exactly, does that look like? Here are three organizational strategies that can help build confidence in an era of data-driven disruption.

Make data education a company-wide initiative

Data literacy is no longer the exclusive purview of number-crunchers and policy wonks. Given the ubiquity of information in organizations today, and the ever-expanding roster of tools (sanctioned and unsanctioned) employees can feed with it, educating people on the fundamentals of how and when to safely use information is “not only necessary right now, but urgent,” according to Carlos Perez Chalico, EY Canada’s cybersecurity and privacy leader. 

Why? Think of the potential implications of an employee using a dubious AI note-taker to record and summarize a confidential client call, for instance, or of an associate misinterpreting the permissions to use a set of numbers about patient health. “The effects of these issues can be significantly reduced when organizations devote time to and invest in data literacy and cybersecurity awareness,” Chalico explains. That doesn’t mean everyone has to become a statistician, he adds, but organizations should ensure everyone on the payroll has at least a foundational understanding of why it’s important to use data in authorized ways, and what risks failing to do so can create. 

In 2021, Telus launched a company-wide data literacy program to give all employees—regardless of position—a better understanding of data, its governance and its implications for risk and privacy. The program centres on a core curriculum that covers the basics, but is delivered in different ways to better relate the material to the experience and role of each student. “If we are going to be a data-driven organization, it’s important for everyone to have that knowledge about what’s right, what’s wrong, and what’s safe to do, so we can all make more informed decisions,” Snively explains. “That knowledge can’t just live in the privacy office, or the data governance office, or the security office.” 

Data literacy at Telus has improved by 24 per cent across the organization since the program launched, which is changing how employees engage with information and technology. When ChatGPT and its generative AI ilk exploded into the mainstream two years ago, for instance, Snively’s office was immediately flooded with inquiries from staffers about its safety and appropriate use—the kind of good, thoughtful questions that probably wouldn’t have emerged before the company’s education campaign.

Appoint internal ambassadors

The ascent of AI is sparking some to view data governance through a wider lens. “Companies can no longer deal with this in silos,” says Bojana Bellamy, the U.K.-based president of Hunton Andrews Kurth LLP’s Centre for Information Policy Leadership, which has published research quantifying the benefits of a holistic approach to managing data. Because data is now the remit of virtually all facets of an organization, and because there are now so many regulatory imperatives to consider (related to privacy, security, legal and competition, among others), it’s no longer effective—nor particularly feasible—for responsible use to fall to compliance teams alone. “We believe very much in organizational accountability,” Bellamy says. “Companies in this new data world have to be responsible for data that they use, that they create, that they collect, and for technologies that they develop to use it. That goes beyond just complying with the law,” she says. 

Shared accountability manifests in several ways at Telus. The most visible is the company’s cross-functional army of more than 500 employee data stewards— representing every business unit in the organization—each of whom undergoes certification and participates in monthly updates on relevant developments in data governance. Their role is part ambassadorial (in that they disseminate knowledge and new developments to their teams), part ideative (in that they identify opportunities to better use information), and part protective (in that they are trained to identify potentially unsafe or untrustworthy applications of data by their colleagues). 

According to Jesslyn Dymond, director of AI governance and data ethics at Telus, data stewards help operationalize knowledge and judgment that can otherwise seem too abstract or impenetrable for teams to apply. “Our data stewardship program is such an important resource in embedding data governance throughout our organization,” she says. “It means it isn’t just one team’s job to make sure that data can be used, or that we have the right policies and programs in place.”

Put guardrails in place

Machine learning and AI builders are familiar with the concept of guardrails: the mix of rules, tools, and screens that buffer a model’s inputs and outputs, in an effort to produce safer and more accurate results. Done well, they bake in a degree of security without compromising agility.  

The same concept can be applied to organizational processes surrounding not just AI, but any emerging data-based technology. While at first glance operational guardrails like stress tests, risk-management scans and ethical reviews might seem to slow things down, in practice they often serve as business accelerants, because issues tend to be identified and mitigated before they become problems. “When you embed cultures and behaviours of accountability into the DNA of a company, it gives you safeguards to move faster and to create more benefits from the data you have, because everyone knows how to use it well,” Bellamy explains. “Ultimately, you will become a quicker, better business. You will have more customers, and you will earn more trust.”

AI holds tremendous potential to advance dozens of critical applications within Telus, and the competitive landscape doesn’t afford it much time to dither. Yet the risks are too significant to charge forward full tilt. “You don’t fail small with AI,” says Dymond. “We need to not only understand what we can do, but also really think critically about what we should do.” 

Telus was an early signatory to the Government of Canada’s Voluntary Code of Conduct on the Responsible Development and Management of Advanced Generative AI Systems, which recommends (among other things) the adversarial testing of AI systems to seek out weaknesses and flaws—a process known as red teaming. Telus has built on that to apply a technique called purple teaming. “It brings together people who are thinking about how to break the system with blue team individuals who are trying to build solutions,” explains Dymond. The process involves trained developers, of course, but also data stewards and other volunteers from across the organization, so everyone from financial analysts to call centre associates get a chance to stress-test systems before they go live. “You get a real diversity of perspective, and it really enriches the user experience to robustly ensure the product is going to work in the way it is intended,” Dymond adds. And because the various parties work on this collaboratively and simultaneously, this diligence tends to allow the company to iterate and innovate at a competitive pace.

A competitive advantage

In today’s unpredictable techscape, no one initiative can guarantee an organization’s ability to use data responsibly. But together, knowledge- and accountability-boosting actions can establish the kind of reliable rigour that—over time—bolsters customer confidence. “If you think about responsible data practices as drivers of customer trust, you start to recognize the business imperative,” Snively explains. “And if you think of them as something separate from the value of the business—as boxes to check off as quickly and inexpensively as possible—you will never get where you need to go.”

This content was paid for and directed by TELUS and was produced independently of The Logic’s newsroom in consultation with the advertiser. You can read our policies on advertising, sponsorships and partnerships here.