TikTok And Instagram Have An AI Influencer Problem. Blockchain Can Help.

The rise of truly powerful and accessible, Generative AI (GenAI) is beginning to shake up the social media landscape. New AI tools and even influencers are beginning to shape how we interact through platforms like TikTok and Instagram. This is not only troubling in terms of what content can be generated or spread but is also an existential threat to human creators as well. The solution to this problem lies in what blockchain technology can offer in terms of verified, human digital identities.

The Rise Of The AI Influencer

Just recently, TikTok announced it was testing a new in-app tool to allow users to utilize AI to generate avatar images. The tool will let users submit photos of themselves and create unique images in a range of poses and styles. These can then be downloaded, shared, or used as a custom profile picture.

This straightforward tool can enable anyone to have a visually interesting avatar that they may otherwise not be able to create themselves. Furthermore, it’s free and nearly instantaneous, empowering creativity and lowering the barrier to access. That is a noble idea and solid in theory, but it is also just the beginning. Another recent revelation from TikTok is that the platform is also exploring creating AI-driven influencers who could promote brands, talk about various topics, and generally do exactly what influencers already do — only they wouldn’t be real.

AI influencers aren’t entirely new, with prominent profiles like Imma and Shudu attracting hundreds of thousands of followers. However, until very recently digitally created influencers cost more than a real human being to deploy. That’s changing fast. Tools like ChatGPT, Midjourney, and many others are making it affordable for anyone to create high-quality content with just a few sentences, and there’s no reason to think this trend will slow down.

Soon, it isn’t going to be just big brand names who may create their own AI influencers; it will be anyone. Once built, they can then use these personas to promote anything they want. Think about what fake accounts and bots are already doing to social media and it shouldn’t be hard to imagine how dangerous this could be.

In the words of Ed Keller, a leading authority on word of mouth and influencers, and author of the Influentials, “At the core of word of mouth’s power, including word of mouth from creators, is trust. That’s what makes it so powerful and is fueling its growth. The influence of unchecked, unverified AI is troubling, and shows how quickly new technology might destroy trust in the online world. This is bad news for everyone who uses the internet, and particularly for real-life influencers because they have such a big digital footprint that can be used to train hyper-realistic, ultra-believable clones of themselves. There are potential negative consequences for marketers as well, if the rise of fake creators undermines the very foundation that has led to the huge rise in the size and credibility of the creator economy.”

The threat of bad faith actors spreading misinformation and hateful content, influencing public opinion, and even potentially shaping the course of politics is too real to be ignored. At this point, there is no putting the genie back in the bottle. As OpenAI made clear in a recent statement addressing criticism from Elon Musk: “We’re making our technology broadly usable in ways that empower people and improve their daily lives, including via open-source contributions. We provide broad access to today’s most powerful AI, including a free version that hundreds of millions of people use every day.”

Even if a given government tried to combat this new online reality, by creating rules around what type of AI content can be created and spread, they wouldn’t be universally applied. Additionally, some people will always be willing to break the law, so while solid regulations should be put in place, this problem is bigger than what the legal world alone can tackle.

How Blockchain Technology & Digital Identities Can Save Us From The Machines

We have to accept that there’s no going back, while also designing a practical solution for moving forward. This is where blockchain can support us. Instead of controlling what is allowed to be generated, a system can be created to identify the content’s original source. Whether it be a human or an AI, consumers will be able to see where it came from and make informed decisions about how they interact with it.

This is especially important given the importance social media has on society. Marketing strategist and best selling author Mark Schaefer believes we’re on the cusp of a profound moment in business. “From now, on, the number one question our customers will have is, ‘Is this real?’ Businesses must spare no expense protecting their brand, executives, and employees from deep fakes. I think blockchain has to play a role in this as a source of truth.

“I’m concerned that businesses have no idea what’s coming,” Schaefer said. “A brand is trust. No trust, no brand. That should worry everyone.”

Recently, leaders in Davos were thinking about how blockchain can be used to track, sort, and manage the data that AIs are being trained on. As Sheila Warren, the CEO of the Crypto Council for Innovation, stated to CNBC: “I actually do think that the verification of an AI and sort of the checks and balances …are going to be blockchain driven and blockchain-backed.” This is a good start for building trustworthy autonomous identities, but what about untrustworthy ones?

Decentralized technology can provide credentials in a fully verifiable and unfalsifiable way, due to the immutability and transparency of these networks. Blockchains cannot realistically be hacked or manipulated, and any changes made to them are always able to be seen by all participants.

If we want digital profiles that come with verifying credentials, this technology is the way to go. With this, Digital IDs can be provided to every human and artificial participant on social media, defining their online persona, history, preferences, and any other information they choose to share. In the interest of privacy, much of this data would be self-encrypted and under the control of the users, who could decide what is and isn’t for the public to see. However, one parameter that should always remain public is their proof of humanity or lack thereof.

Real people can perform a “humanity check,” to prove they are real, and human influencers can additionally link a wide array of credentials to make it clear they are who they say. However, any AI influencers would also need to be transparent about that fact, in a way that cannot be circumvented.

In such a system, any profile that doesn’t have the humanity credential could be assumed to be machine-generated. This doesn’t mean this content is nefarious, but it would give consumers a degree of awareness of what they are taking in. Social media platforms would be free to tag, block, or filter AI content in any way they saw fit, but gone would be the days where bots could influence the social discourse without oversight.

This isn’t like the “blue check” on X, where users pay a fee and get an icon next to their name and that’s about it. This would be available exclusively through the user providing some form of meaningful proof of who they are, probably confirmed by another verified human, and only then would their blockchain-enforced credential be awarded. With this, users could trust that a verified account is exactly that, and not just someone who was able to pay the entrance fee.

Concluding Thoughts on AI Avatars

If built correctly, such a system could allow us to navigate this brave new world of AI in a way that retains individual privacy, the freedom to choose human sources, and the ability for people to make a living via content creation. Along the way, there will be many details to navigate, but there is currently no other plausible model for protecting social media from the growing impacts of AI. Laws can only go so far, we need something practical to put in place.

Blockchain and digital identities provide a realistic means to protect consumers and the human workforce. This is a step we need to take soon because the speed of AI development isn’t slowing down. However, there is an opportunity for a more powerful and safer internet for the next generation and beyond.