AI-created bots, social media and advertising

Benjamin Barrier at DataDome asks what Meta’s bot experiment means for social media and advertisers

 

Meta opened 2025 with backlash from users in response to increasing media attention around the company’s use of generative AI to create bot accounts. Users dug up historical bot accounts and brought up questions around the platform’s non-transparent use of bots, citing issues around cultural appropriation, sloppy imagery, and tendency to lie in chats with human users.

 

The backlash of the past month has raised questions around the implications of bots on social media. Now the floodgates have opened, we’re likely to see another, more foolhardy, social platform look to institutionalise bots in a similar way. What will this mean for users, social media platforms, and the advertising industry?

 

Digital advertising’s bot dilemma

While the social media behemoth has now deleted all historical bot accounts, the platform’s (now abandoned) plans to roll out more AI products and potentially new bot profiles should worry not only social media users but also digital advertisers.

 

If bot profiles gain a larger presence on the platform, we’re likely to see a breakdown in users’ trust in Meta and other social media companies. Users won’t have the same confidence of knowing if they are reacting with a real human user or an artificial account, and might look to other platforms for a more authentic experience.

 

But the problem of trust for digital advertisers goes beyond just users’ trust in social media platforms. The budget assigned to social media advertising each year is huge: in 2024, global advertising on social media was estimated at $234.14 billion, a 140% increase from five years earlier. 

 

Digital advertisers have historically been willing to splash the cash on expensive social media campaigns despite concerns about bot-driven ad fraud and its potential effect on ROI. Click fraud is one increasingly common example, with armies of bots targeting pay-per-click (PPC) ads on social media and elsewhere. Having hit their ‘click limit’, companies are then forced to spend more money to place ads, but this money is really just going into fraudsters’ pockets.

 

With bot traffic already accounting for almost half of all web traffic, it’s not hard to see why marketing teams are sceptical of the ROI offered by digital advertising. And with the introduction of bot personas on platforms like Meta, this problem is only set to grow.

 

Unless Meta makes a point of separating its own bot traffic out from paid traffic, and signposting this clearly to advertisers, it’s inevitable that decision-makers will be concerned about wasting ad spend on synthetic users who will never convert to customers.

 

By encouraging bots onto social media platforms, it’ll also become harder for marketing teams to track the engagement around organic content. With approved bot accounts (at least in Meta’s case) able to interact with posts by liking or commenting, marketing teams will no longer be able to rely on their usual engagement metrics like impressions and clicks to tell them if a post is performing well among real users.

 

Again, Meta will have to go out of its way to be as transparent as possible if they want to stamp out these concerns, for instance giving decision makers the option to remove bot engagement metrics when looking at the total metrics for a post.

 

Beyond advertising

Outside of the digital advertising industry, the introduction of bot personas on social platforms is likely to influence public discourse around major news events and could even lead to misinformation.

 

We’ve seen this happen repeatedly via the medium of non-official bots on social media - for instance bots were accused of spreading ‘fake news’ during the COVID-19 pandemic and shaping the narrative over the course of the 2024 elections.

 

The introduction of artificial ‘bot influencers’ on social platforms threatens the role of the real influencer, an industry that stands at $21.1 billion (as of 2023). We’ve already seen mixed reactions to AI bot influencers like Miquela and others, and official bot accounts on platforms like Meta suggest this part of the platform could expand further.

 

And because authorised bot accounts will be able to like or comment on users’ posts, it’ll become even harder to identify if an influencer is trending because of their own merit - or because the social platform wants them to.

 

What’s next for social media bots?

It’s unclear where social media bots will go from here. While Meta might be wary of experimenting with bot features again in the near future, it’s possible the whole discourse will have inspired other social platforms to follow suit, not least as a way of attracting media attention. 

 

As we navigate a world in which bots are not only tolerated but encouraged and authorised, social media companies will need to promote greater transparency about their bot policies. We’ll also need better regulation in place that requires companies to report on bot activity levels - otherwise we could see social media platforms become a Wild West for bots.

 

---

 

Benjamin Barrier is co-founder and Chief Strategy Officer at DataDome

 

Main image courtesy of iStockPhoto.com and georgeclerk