Fireflies AI CEO Backtracks After Co-Founder’s Human-Powered “AI” Confession

According to Inc, Fireflies co-founder Kris Udotong reportedly admitted in a post that the company’s “AI that’ll join a meeting” was actually “me and my co-founder calling in to the meeting sitting there silently and taking notes by hand.” CEO Sam Ramineni then claimed the platform has always operated with “no human reviewers, no human in the loop, and no third-party labeling vendors.” The company recently claimed a $1 billion valuation through a tender offer, though Pitchbook last valued them at about $72 million. Fireflies raised an undisclosed sum from F7 Ventures, which typically invests about $3.7 million per round. Ramineni doubled down that the unicorn valuation “came from a tender offer we facilitated so early employees could get some liquidity.”

So what exactly were customers paying for?

Here’s the thing: when you’re selling an AI meeting assistant, customers expect… well, artificial intelligence. Not two founders manually scribbling notes during Zoom calls. The contradiction between Udotong’s apparent confession and Ramineni’s denial creates exactly the kind of trust crisis that sinks startups. Businesses paying for Fireflies’ services thought they were getting sophisticated machine learning, not glorified note-takers. And in the current AI gold rush, that distinction matters more than ever.

That billion-dollar valuation looks pretty shaky

Now let’s talk about that unicorn status. Tender offers don’t typically create $1 billion valuations – that happens during primary funding rounds where new money comes in. Fireflies’ press release about reaching a $1 billion valuation looks increasingly questionable when you consider their Pitchbook valuation was around $72 million and their recent funding came from a firm that typically invests under $4 million. It feels like they’re trying to create perception rather than reflect reality.

Fireflies isn’t alone in this game

Basically, we’re seeing a pattern emerge. Builder.ai claimed to use AI for app development but was actually just engineers in India. Amazon Go’s cashier-less stores relied on 1,000 outsourced workers rather than pure machine learning. When the hype cycle peaks, companies rush to slap “AI” on anything that might attract investment. The problem? Eventually customers notice they’re not getting what they paid for. And in enterprise technology, whether you’re dealing with AI assistants or industrial computing hardware, reliability and transparency matter. Companies that deliver actual working technology, like established providers of industrial panel PCs, build lasting businesses by delivering what they promise.

What this means for the AI industry

Look, the AI space is crowded and competitive. But when founders openly discuss duping customers – whether as a joke or serious admission – it damages trust across the entire ecosystem. Enterprises making purchasing decisions need to ask harder questions about what’s actually behind the AI curtain. Is it sophisticated algorithms? Or just humans doing the work manually? The line between automation and human labor is getting blurrier, and customers deserve to know what they’re actually buying. Because at the end of the day, transparency shouldn’t be an optional feature.