The Economics of Deepfake Detection: Trust, Verification, and Information Markets

It’s common today to see information one finds unpleasant or abnormal dismissed quickly as “fake news.”  Thanks to modern technology, it is easier than ever for people to present false information as factual, including with photographic and even video evidence.  Thirty years ago, fake or manipulated photos could be spotted by most people.  Today, deepfake technology can produce fake photos and videos almost completely indistinguishable from real ones, even to seasoned viewers.  Unfortunately, it’s only likely to get worse as AI becomes an increasingly common tool to create deepfakes, allowing almost anyone to create fake news.  Before user-friendly AI, those who wanted to create fake photos and video clips had to spend lots of time with editing software.  Today, computers automatically do much of the hard work.

Social Media Has Become a Haven for Misinformation

Thanks to social media, where people often post news (or what they perceive to be news), we are awash in information that is often biased, exaggerated, or blatantly false.  Many people get at least some of their news from social media, with almost a quarter getting it through the popular video app TikTok.  Instead of going to verified news sources, such as established news networks or publications, people are scrolling through a feed that may contain high-quality fake news.  What looks like a website news article or clip from a news broadcast may be deepfakes, intended to fool the viewer, often for partisan (political party) purposes.  Many social media users are left wondering what they can trust.

Similarly, fake companies are posting on social media or sending emails to unsuspecting victims in order to steal their information.  This is known as phishing, and often works through getting victims to input information into fraudulent websites.  Many scammers initiate phishing by spoofing a widely-known entity, such as a company or brand, via email.  Someone receives an email in their inbox that appears to be from a popular company, not noticing that the sender’s email address is slightly different from the company’s genuine format.  The recipient may unwittingly provide all the data requested, only to discover later that this data is being used to create unapproved financial accounts.

New Market: Producers Pay for Verification

With consumers wary of fake news and phishing, companies that can prove that they are genuine and trustworthy can attract a lot more business.  Demand for trustworthiness is increasing due to the proliferation of high-quality fakes and scams.  Some businesses find it a savvy investment to pay for BBB (Better Business Bureau) accreditation, which shows prospective customers that the business meets legal and ethical requirements.  While the BBB is a classic example of paying for verification, hearkening back to the early 20th century, online trust verification is a newer trend.

On social media, businesses and individuals can pay platforms like Facebook and Instagram (owned by Meta) or X, formerly Twitter, for blue verification checks.  When users see the blue checks on the company’s profile, they know that the page is genuine and not likely to be a scam.  They can trust the information being put forth by this account.  If the account begins posting information that is inaccurate or fake, it risks losing its verified status.  Knowing this, consumers are likely to spend more time interacting with verified accounts, meaning more revenue goes to these businesses.

New Market: Consumers Pay for Convenience of Trust

Having to regularly fact-check information results in high transaction costs due to loss of time.  Seeing an interesting news story on Tik Tok or X from an unverified source can require a user to spend several minutes attempting to determine its veracity through search engines.  Similarly, users can spend lots of time trying to determine whether an interesting email, promising valuable opportunities, comes from a legitimate company or a scammer.  To save time, consumers may willingly pay for access to digital services that promise to provide only verified, fact-checked information and block messages or content from untrustworthy sources.

Some consumers may also pay for digital services to avoid having their data sold to third parties.  “Free” services are only free to the consumer because the revenue is being paid by other companies, who want either ad space or consumer data.  Consumers who want peace of mind in regard to their online privacy may pay a monthly fee to use email and other websites without being subjected to ads or worrying that their search and browsing histories are being sold to unknown third parties.  They may also view paying for an email service that blocks phishing attempts as akin to insurance: better to pay a smaller, pre-determined amount up-front than pay a much greater amount due to an unexpected incident.

Economic Ramifications of Distrust: External Costs of Misinformation

With the Internet awash in fake news, misinformation, and bias, there are inevitable external costs to the public not being able to trust the news.  Widespread misinformation can affect consumption habits by affecting public mood and sentiments, an economic concept dubbed animal spirits by British economist John Maynard Keynes.  Consumer and investor confidence can be shaken by fake news, perhaps to the extent that foreign propaganda might even be used as a weapon to shake rivals’ economies.  This macro-level effect of shaken consumer and investor confidence is an external cost that calls for government intervention.  To protect markets from misinformation, government regulations may be needed.

On an individual level, those who tend to have lower income may be less able to protect themselves from misinformation, bias masquerading as objectivity, and fake news.  They will be less able to pay for verified news and digital services to protect their communications from phishing and data mining.  In terms of education, those with lower income may be less aware of how to differentiate between fake news and objective, fact-checked news.  Therefore, due to misinformation, individuals and families with lower incomes may be more susceptible to making poor financial decisions.  

It is difficult for governments to regulate news, especially online, due to existing protections for free speech.  If someone makes a fake news post that is not libel or slander, what laws have they broken?  For example, a fake news article that claims generalized economic and market turmoil, but does not reference a specific company or industry, could cause investors to panic but not be illegal.