By: Praful Krishna

Cambridge Analytica, the political consulting firm that worked for the Trump campaign and harvested raw data from up to 87 million Facebook profiles, has folded. In the wake of Facebook’s congressional hearings and the role Cambridge Analytica played in the Trump’s election campaign, considerable concerns have surfaced. 

Facebook has long been regarded as the trailblazer in social media. However, how much can users really trust platforms like Facebook with their data? In this case, a third-party developer exploited a loophole in the system to gather information on users—as well as users’ friends—without them knowing. Facebook knew—and did nothing.

Zuckerberg faced one of his most stringent tests in the uncomfortable questions by senators on Capitol Hill. The job of CEOs and CIOs could become even harder if the U.S. decides to follow the lead of the European Union, which is set to embrace the EU General Data Protection Regulation (GDPR) that went into effect on May 25, 2018. 

This episode comes at a time when CIOs across sectors, especially service providers, are thinking of adopting AI. There are profound lessons for everyone to study.

The rise of AI as the New Masters of Data

Ten years ago, social media was in its adoption curve where AI is today. Large companies are rapidly embracing the idea, and only opening to adoption now; a study by Adobe says only 15 percent of enterprises are actively using AI, while 31 percent are expected to add it in some form or another over the coming year. As most of its breakthrough abilities continue to unfurl in the fields of finance, healthcare, agriculture, manufacturing, technology, and countless other areas, an overlooked detail is that data fuels this technology.

In the coming decade, access to data to train AI and its wise usage will create winners and losers in the business world. As data becomes the new valuable commodity replacing oil, it’s natural that the companies developing AI products will be eager to improve their AI engines to offer better products and services. Moreover, this can happen only when they get more and more data from their clients.

There are nascent algorithms like Calibrated Quantum Mesh, or applications like cognitive automation, which differentiate in that they need very little data to train. However, it is still meaningful. Additionally, even if they don’t need data to train, they process enterprise data—or have access to it.

In other words, AI product companies are potentially getting data out of their client firewalls and exposing it to all kinds of risk. Enterprise data is, in most cases, customers’ data metamorphosed. It took various high-profile data breaches for organizations to stop working under the assumption of “if,” and build strategies around “when” a data breach will occur. Similar is the case for ethics when it comes to customer data. When customers register to avail any new service or product from a business, they put their faith in that business to protect their data.  It’s the responsibility of enterprises to protect that data at all costs.

A new study from Accenture shows that the scope of digital risk has expanded beyond cybersecurity and privacy into what is coined as digital ethics. Enterprises can’t deny their responsibility for what they do with the customer data. Yes, it’s about ethics now—how enterprises act on the data they collect and analyze has come under the scanner. The study says that in the future digital economy, enterprises and government agencies that achieve the proper balance of managing security risks and building digital trust effectively will thrive.

It’s critical that enterprises integrate ethical data practices throughout their business processes with more robust ethical controls.

AI Product or AI Solution – The Nuance Matters

CIOs need to be careful. Let's say that, with appropriate diligence, they are able to find trustworthy product vendors that can minimize security breaches. Still, they would have given their most important competitive advantage—data—for the benefit of the product vendor and, by implication, their own competitors. The idea that started in an attempt to strengthen the company’s competitive differentiation may end up achieving just the opposite.