AI’s Digital Gold Rush Is Fueling a Trust Crisis—And No One Is Ready for the Fallout

In the meta-narratives around generative artificial intelligence, Chat GPT, Gemini, Copilot and Apple Intelligence are either powerful tools that are leading us to a brave new world of content creation and interaction or to humanities’ doom. 

The problem with Panglossian or dystopian narratives is they fly over nuance. The reality is that both things can be true at the same time. The yellow brick road is paved with gold, but the destination may be smoke and mirrors where our ability to tell reality from fiction is lost.  

When it comes to digital content, this has real consequences. If consumers don’t know what they can trust, there will be no Hollywood ending to the digital journey we are on. 

Trust is the foundation of the digital economy. Consumer decisions about the credibility of content, be it news, social media or emails from friends, purchase decisions or the payment of invoices is all based on the belief that what they see is real.  

The economy has always been based on trust, but it’s being steadily eroded. The rise of generative AI, with its shiny and largely unthinking adoption path, is turbocharging the rise of fake content, video and criminal fraud. The consequences are real and expensive. 

As an example, a business leader requested an invoice from an agency she was working with. Not knowing her email had been compromised, cybercriminals sent a fake payment request from what looked like the agency email. Her accounting team paid the invoice for more than $250,000 to the bank details that were provided, only to subsequently receive a call from the agency that they had not set a bill. A quarter-of-a-million-dollar loss is a serious wake-up call.  

The rise in fraud on platforms like Facebook Marketplace, where transactions may be smaller, but the breadth of criminal activity is much broader, is furthermore a major concern. When every victim loses a little bit of faith in ecommerce, the foundation of trust is eroded. For every AI-generated YouTube click-bait video that purports to show “almost” impossible to imagine images, fake Joe Biden robocalls, a drunk vice president, the ability to determine what is and is not real is corroded. 

The technology trifecta of personhood verification, content provenance authentication and creation of safe spaces to build trust in content, has not received the same attention as the rise of generative AI. But it should.  

Personhood Verification 

User identity verification is far from new, but its adoption is increasing. DocuSign for many years has been able to provide a high degree of confidence in the digital identity of signers to meet the legal threshold for contracts.   

We’ve moved a long way from Twitter’s pay to-be-verified blue checks to more sophisticated verification systems. LinkedIn’s use of CLEAR to verify that users are who they say they are, using a government ID or data, provides a high degree of confidence in user identification. 

Is it possible to fake verification? Yes. But the barriers are clearly higher. Combined with widely used two-factor authentication for bank or other digital accounts, personal identity verification may be imperfect, but those who are not using it are more likely to be the easier target for fraudsters.  

While identity verification solves the problem of the ability to trust the identity of the person or entity with whom I am engaging, it doesn’t provide a mechanism for determining if content was produced by them. 

Content Provenance Authentication 

Verification and authentication are used somewhat interchangeably. At Tauth.io, we are focused on one of the most significant digital trends since the Internet was founded – the ability to authenticate the provenance of digital content. 

There is a tremendous amount of work going on in this space largely-driven by the Content Authenticity initiative (CAI). The world’s largest technology companies and content generators are members. The CAI initiative is supported by Adobe, which is working to integrate authentication of the publisher of a document into PDFs, and build an ecosystem around CAI principles.  

Tauth.io is a CAI member. It’s one of a number of technology companies building authentication solutions focused on building trust around the provenance of the documents to help consumers know they are authentic. To achieve this goal, we use a simple user content authentication process that uses C2PA open-source protocols and Public Key Infrastructure (PKI) to create a robust, irrevocable digital signature based on the confirmation of the unique identity of the author. This is carried in its metadata in the form of a digital watermark. Once a document has been authenticated by the publisher it is logged using a blockchain hash. 

The key to the success of this process is simplicity for the end user. An application will enable one-click signing and authentication. When a user opens the document, a browser plug-in recognizes that the document has been authenticated and is trustworthy. The digital watermark provides the link to the blockchain record for the user to determine its author and changes that may have been made. If a document is not authenticated, users are flagged so they can take additional steps to determine authenticity.  

Using this approach - whether or not the blockchain link is included – builds in the accountability of a name attached to a document as the basis for trust. 

Once a document has been signed, the metadata provides a basis for a range of tools to log copyright and identify unsanctioned or fraudulent uses of documents. The approach is well-suited to specific applications and use cases, but not one-size fits all. It should be seen as an important part of the technology solutions to maintain trust in digital commerce.      

Digital Safe Spaces                                                 

It’s important to note that the approaches being adopted by social media platforms for user verification reflect efforts to create safer spaces for those users. However, the business dynamics of social media favor mass audiences rather than narrow, focused ones. Today that has generally meant that verification has taken a back seat to getting as many people on the platforms as possible.  

This creates an opening for more narrowly focused platforms to lean into creating safe spaces where all the users are verified as a strategic advantage. Media.com is one such platform with big ambitions of creating a community of verified users in which content can be trusted and is more likely to resonate.  

While we expect to see more safe spaces being set up for issue-focused communities, the challenge is the reverse of the large platforms. Over the short-term, although the ability to scale is limited by the need to have verified users. Over time, as recognition grows around the importance of user-verification, technology solutions will help simplify this and drive adoption. For both existing platforms and new communities, verification is likely to become table stakes. 

User verification, the authentication of content provenance, and safe-space communities all have a critical role to play if we are to maintain a foundation of trust in the digital economy. As AI-generated content plays a growing role in the ecosystem, and the issue of fake and fraudulent content becomes increasingly problematic, we will see growing focus and innovation on the other side of the equation - digital safety. 

Developments on both sides will be synergistic and are likely to maintain the system’s balance. Ultimately, we believe this will ensure that the digital economy will not grind to a halt, but continue to thrive.