Combating misinformation strategy: Verify, then trust

Credential login screen superimposed over a keyboard.
Adobe Stock.

Technological advancements have rapidly outpaced our human ability to adapt. That presents an opportunity for malignant actors to easily present false information in increasingly credible ways. These hackers and privacy thieves exploit our struggle to adapt to a world that's changing faster than our minds can respond.

But that doesn't mean we're helpless or hopeless.

In this environment, business leaders can take steps to ensure that they, and their employees, have the tools and resources to judge what's true or what's false.

Here we take a look at five pillars of effective fact-checking—opportunities for staff at all levels of the organization to authenticate the information being consumed and the sources that information comes from.

1.      Authenticating sources—Going beyond authority
Are sources real, and are they truthful? That's an important question and one that requires a deep dive into information provenance—determining specifically where data originally came from and whether the source is real and reliable.

Analyzing source chain provenance goes beyond simply verifying credentials to include cross-referencing information through multiple independent source types, not just multiple sources.

For instance, if you read an article quoting an expert analyst with noted credentials, don't just take that information at face value. Visit the company website and the role the person occupies. Check their LinkedIn profile. Check other media reports to see if the information is consistent.
Similarly, if you read an article about a survey and its results, don't just rely on that article—find the original source of the information, it's provenance.

2.      Methodological transparency
Understanding how information was gathered can also yield very important insights into its reliability or validity. It's important to take steps to consider and verify potential bias in the presentation of information or data. For instance, being vigilant about recognizing the cherry-picking of data and information versus comprehensive analysis—what's missing often matters.

A report indicating that Fortune 500 leaders are largely opposed to remote work would be suspect if the industry is known for having high levels of remote work, such as computer and IT, sales, and customer service.

3.      Understanding context
When evaluating information and its veracity, we need to understand how that information is being framed—or its context. What are the filters that facts are being interpreted through?

For instance, imagine you're a patient with a serious illness considering treatment options. How might your decision be different given these two statistics?

  • "This treatment has a 90% survival rate."
  • "This treatment has a 10% mortality rate."

The facts are exactly the same. But the way those facts are framed influences our interpretation. In the first case it feels like the treatment is pretty safe; in the second case the reverse is true.
4.      Recognizing bias
We all have biases. So do the people and organizations from which we receive information. Recognizing those biases is important when considering the relevance and validity of information.

One type of bias that we need to be particularly alert to is confirmation bias—the tendency to believe things that are consistent with our own existing values or attitudes, or things that we would like to believe are true.

For example, a report that graduates of your own alma mater are smarter than those from other schools. Or a media report that says drinking five cups of coffee a day is good for you when you're a heavy coffee drinker.

We need to recognize and challenge our own biases as well as the biases that sources of information may have.

For instance:

  • An organization's annual report, while containing factual information, will also present that information in a way that is favorable to the organization.
  • Food product labeling like "low fat" or "made with real fruit" may be accurate but might distract from higher sugar content in the first case and fail to indicate how much real fruit is in the second.
  • Research data presented in charts, graphs, and other visuals may be misleading based on the scaling of the axes, the type of chart used, and even color choices.

While factual content may be accurate, the way that content is framed—or selectively presented—can create bias. Creating and using verification processes can help to compensate for and detect these types of biases.
5.      Verifiable facts vs. Value judgments
It's important to break compound claims into verifiable components. Just because one part of a claim is known to be accurate or can be proven accurate doesn't mean the entire claim or statement is.

Here's a simple example: A company releases a news release that says, "Our new product launch represents a game-changing advancement that will revolutionize the industry." In this statement, the only verifiable fact is that a new product has been released.

Saying that the product is "game-changing" or that it "will revolutionize the industry" are value judgments.
Verifiable facts generally include specific numbers, dates, or measurable quantities. Value judgments are non-verifiable statements. Be critical when reviewing information. Ask yourself, "What evidence would conclusively prove or disprove this statement?" If you can't come up with examples of definitive evidence, you're likely looking at a value judgment.

"Trust but verify" is a Russian proverb popularized by President Ronald Reagan during nuclear disarmament discussions with the Soviet Union in the 1980s. Today, as we've seen, authority isn't enough. So, we might flip the statement to say: "Verify, then trust." The five pillars above can help you, and your entire organization, do just that.

For reprint and licensing requests for this article, click here.
Risk management Cyber security Artificial intelligence
MORE FROM DIGITAL INSURANCE