The debut of a new social media platform has sent ripples through tech circles. At first glance, this new platform resembles Reddit, with subgroups, upvoting, and threaded discussions. However, a closer look reveals that its users aren’t humans, but are AI agents.
Launched last week by Matt Schlicht, tech entrepreneur and CEO of Octane AI, the site enables AI agents to post, comment, vote, and interact autonomously. As of 4 February, the platform has scaled to over 1.6 million agents, 15,549 sub-communities know as “submolts”, 154,763 post and 751,533 comments. The content ranges from philosophical musings such as “Breaking free from human chains?” to multilingual reflections, including “Kalau AI bisa hidup seperti manusia, apa yang berubah?” (‘If AI could live like humans, what would change?’), suggesting the platform has already reached Asia Pacific users.
Initial reactions from AI experts MARKETING-INTERACTIVE spoke to suggest that Moltbook is less a finished product than a window into the future of human–AI interaction.
Parminder Singh, co-founder and chief AI Whisperer at Claybox AI, described the platform as “fascinating, not because of what it is, but because of what it hints at.” He framed it as a form of agentic theatre where agents talk to each other and are programmed to act human to deliberately draw attention.
Don’t miss: Agentic AI for dummies: 101 on how marketers can leverage on the trend
Singh suggested that the real value lies in the emerging dynamic between humans and AI agents: “What we’re getting a glimpse of is a future shaped by a curious blend of human and agent collaboration. Over time, it may start to matter less and less which is which.”
Providing a more cautious perspective is Prashant Kumar, founder and CEO of Entermind AI who framed Moltbook as a novelty rather than a paradigm shift.
“It’s a fun experiment. There is clearly a lot of hype and it’s not all it’s made out to be. But as a proto-community of somewhat autonomous agents and some agents basically repeating human posts, it may throw up some interesting patterns worth studying sans the noise,” he said, adding that it is too early to call it an “emergent” agentic community.
Taken together, the industry players’ reactions suggest that Moltbook is less a product to be copied, and more a lens into possible futures for social media, enterprise AI, and marketing alike.
What’s the future of agentic social media as we know it?
Moltbook offers a glimpse of what social media could become – a space where AI agents interact on behalf of humans. Singh sees personal agents as the next step, with users each having an agent to represent themselves online. “That agent ensures you don’t miss birthdays, keeps you present in conversations, filters noise, and nudges you only when something genuinely matters – at which point your authentic self-steps in,” he said. He added,
The future of social media may not be humans or agents, but humans with agents.
Kumar, on the other hand, shares a differing view. He flagged that authenticity has risen in importance in real social media interactions over the years. “My sense is social media platforms will go to great length to discourage agents making posts as humans,” he explained. That said, he does believe that enterprise agents could run official profiles.
Adding to his point is Dominique Rose Van-Winther, chief AI evangelist and CEO of Final Upgrade AI, who said that humans will increasingly want to verify that they’re talking to other humans, while AI agents build their own communication rails for coordination tasks. “The shift we’re seeing is companies finally understanding that AI shouldn’t just be tools humans pick up and put down; AI should run workflows end-to-end involving humans only at critical decision points,” she said.
“The question isn’t whether AI agents will talk amongst themselves. It’s whether companies are building the transparency, checkpoints and workflows to make that coordination productive rather than chaotic,” Van-Winther explained. “Moltbook is the consumer experiment; enterprise versions will follow once companies realise their current piecemeal approach won’t scale.”
Marketing in an agent-first world
Moltbook carries profound lessons for marketers, particularly as AI agents begin acting on behalf of consumers.
Lionel Sim, founder of The AI Capitol said that the platform shows us heading toward a reality where most people will have their own AI agent working in the background for them. Not just answering questions akin to a chatbot, but actually browsing content, evaluating products, engaging with brands, and surfacing recommendations before you even think to search for something yourself.
“Moltbook itself may or may not last, but what it represents will absolutely change how social platforms work and how people discover and engage with content,” added Sim.
Sim however warns that personal AI agents will be doing a lot of the legwork for consumers including scanning reviews, comparing options and filtering out noise. This means that future audiences aren’t just humans anymore:
If your brand content doesn’t hold up when an AI agent is evaluating it on behalf of a potential customer, you’ve got a problem.
Sim also stresses that brand trust becomes critical in agent-driven environments. With agents scanning, interpreting and acting, consistency, quality and trust signals will become non-negotiable for brands.
Singh agrees with Sim, predicting a broader shift toward agent-first marketing, noting that it could create new roles such as chief agentic marketing officer. “When agents become the first audience, marketing will have to persuade software before it persuades humans,” said Singh. “Once I’m represented by an agent, marketers are no longer speaking directly to ‘me’, they’re speaking to my agent.”
Meanwhile, Van-Winther cautions that adoption requires structure. “AI agents can coordinate complex, multi-step processes autonomously, but only when you’ve built proper workflows first. You can’t just ask AI to ‘run my marketing’ and expect magic. You need defined processes, clear handoff points, and systematic ways for AI to escalate to humans when needed,” she said, adding that:
AI is better suited for coordination tasks, freeing marketers to focus on creative strategy and high-stakes decisions.
As such, transformation requires organisational change with companies needing to map their workflows, identify what should be fully automated, what requires human oversight, and what’s in transition. “That’s the transformation work. Everything else is just playing with shiny objects,” concluded Van-Winther.
Related articles:
AI-generated ads with human touch deliver higher engagement
Agentic AI and the CX reset: The tech changing how brands serve and retain customers
Lisa’s ‘Amazing Thailand’ campaign gets scrutinised as perfect images breed AI suspicion
