Rahul Roy-Chowdhury had one job. As CEO of Grammarly, the company that built its reputation on helping people write better, he sat down for an interview with CNBC’s “Squawk Box” in late March and proceeded to deliver a performance so riddled with corporate jargon and hollow talking points that it became an instant case study in everything critics hate about the current AI hype cycle.
The interview wasn’t a disaster because Roy-Chowdhury said anything technically wrong. It was a disaster because he said almost nothing at all — and did so using the kind of bloated, machine-generated-sounding language that his own product is supposed to fix.
As AppleInsider reported, the Grammarly chief “steps on the same rake over and over” during the segment, offering answers so packed with buzzwords and so devoid of substance that CNBC anchor Andrew Ross Sorkin visibly struggled to extract meaningful information. The publication called it “an embarrassing interview about AI slop” — and that framing stuck.
The Rake, Again and Again
Here’s what happened. Sorkin asked Roy-Chowdhury straightforward questions about Grammarly’s AI strategy, its competitive positioning, and what the product actually does for users now that generative AI has fundamentally changed the writing-assistance market. These are fair questions. They’re the questions any CEO in this space should be prepared to answer with specificity and clarity.
Instead, Roy-Chowdhury offered responses that could have been generated by feeding a press release into ChatGPT and asking it to sound enthusiastic. He talked about “transforming communication” and “empowering” users. He referenced AI capabilities in sweeping, abstract terms. He didn’t give concrete examples. He didn’t cite specific metrics. He didn’t differentiate Grammarly from the dozens of AI writing tools that have flooded the market since OpenAI’s ChatGPT launched in late 2022.
The irony was thick enough to cut. A writing-tool CEO, unable to communicate clearly. A company that sells polish, delivering rust.
Social media pounced. On X, users shared clips from the interview with commentary ranging from bemused to brutal. Several pointed out that Roy-Chowdhury’s answers sounded like they’d been written by the very AI slop — low-quality, AI-generated content — that has become one of the internet’s most persistent complaints. Others noted that Grammarly’s own marketing materials warn against exactly the kind of vague, jargon-heavy communication the CEO was deploying on national television.
One user on X summarized it neatly: “The Grammarly CEO just gave an interview that Grammarly would flag.”
The moment lands differently depending on your vantage point. For consumers, it’s funny. For enterprise customers evaluating whether to spend six or seven figures on Grammarly’s business tier, it raises legitimate questions about whether the company’s leadership understands the product’s value proposition well enough to articulate it under mild pressure.
And make no mistake — Sorkin wasn’t hostile. He was doing his job. The questions were softballs by financial-television standards.
Grammarly has been on an aggressive growth push. The company, founded in 2009 by Ukrainian entrepreneurs Max Lytvyn and Alex Shevchenko, has expanded from a browser-based grammar checker into a full AI writing assistant with enterprise ambitions. It raised $200 million in a 2021 funding round that valued the company at $13 billion. Roy-Chowdhury, who joined as CEO in 2023 after a stint at Google, was brought in specifically to lead the company through its AI transformation and push toward profitability and a potential IPO.
That context makes the interview stumble more consequential than a single bad TV appearance might otherwise be. Roy-Chowdhury isn’t a founder who can fall back on technical genius or origin-story charisma. He’s a professional operator hired to be the face of Grammarly’s next chapter. Communication is literally the job.
The Bigger Problem: AI Companies Can’t Explain What They Do
But this isn’t just a Grammarly story. Roy-Chowdhury’s interview is a symptom of a much wider affliction in the AI industry: executives who cannot describe their products in plain language.
Across the sector, CEO after CEO steps in front of cameras and conference audiences and delivers the same soup of abstraction. “We’re transforming how people work.” “We’re building the future of productivity.” “Our AI understands context.” These phrases mean nothing. They convey no information. They don’t tell a customer, an investor, or a journalist what the product does, how it does it, or why it’s better than alternatives.
The problem has gotten worse as the AI arms race has intensified. Companies that once had clear product identities — Grammarly checks your grammar, Jasper writes marketing copy, Notion organizes your notes — have all converged on the same amorphous “AI-powered everything” positioning. When everyone claims to do everything, nobody can explain what they specifically do well.
This isn’t a trivial communications failure. It has real business consequences. Enterprise buyers are increasingly skeptical of AI tools that promise transformation but deliver incremental improvements. A February 2025 survey by Gartner found that 38% of enterprise technology leaders said they were “disappointed” with the ROI of generative AI investments made in 2024. The top complaint? Vendors couldn’t clearly articulate what problems their tools solved.
Grammarly faces a particularly acute version of this challenge. Its original product — catching typos, suggesting better word choices, flagging passive voice — was easy to understand and easy to demonstrate. You typed a sentence. Grammarly improved it. The value was visible in seconds.
Now the company is trying to sell something far more complex: an AI communication assistant that can draft emails, rewrite documents, adjust tone for different audiences, and integrate with enterprise workflows. That’s a harder pitch. It requires specificity. It requires showing, not telling. And it requires a CEO who can model the kind of clear, effective communication the product promises to deliver.
Roy-Chowdhury didn’t do that. And the internet, which has developed an extremely sensitive nose for AI-generated or AI-influenced corporate speak, called it out immediately.
The term “AI slop” has entered the mainstream vocabulary with remarkable speed. Originally used to describe low-quality AI-generated content flooding social media and search results — think those bizarre AI-generated Facebook images of Jesus made of shrimp, or the SEO spam articles that read like a chatbot having a stroke — the term has expanded to encompass any communication that feels hollow, formulaic, and machine-adjacent. You don’t have to have been literally written by AI to sound like AI slop. You just have to sound like you’ve absorbed so much corporate jargon that your language has lost all human texture.
That’s what happened to Roy-Chowdhury on “Squawk Box.” His answers weren’t wrong. They were empty. And emptiness, in the current media environment, gets punished fast.
The competitive pressure on Grammarly is real and growing. Microsoft has integrated Copilot deeply into Word, Outlook, and Teams — the very applications where Grammarly built its user base. Google’s Gemini is doing the same across Google Workspace. Apple is rolling out its own writing tools through Apple Intelligence. These aren’t startups. They’re platforms with billions of users and the ability to bundle AI writing assistance at no additional cost.
Against that backdrop, Grammarly needs to make a case for why it’s worth paying for separately. That case exists — Grammarly’s tone detection, style-guide enforcement, and cross-platform consistency are genuinely useful features that the big platforms haven’t replicated well. But making that case requires a level of communication precision that the CEO’s CNBC appearance conspicuously lacked.
There’s a lesson here that extends well beyond one company and one interview. As AI tools become commoditized, the companies that survive won’t be the ones with the most impressive models or the biggest training datasets. They’ll be the ones that can explain, clearly and specifically, what they do for users.
That means fewer abstractions. Fewer buzzwords. Fewer claims about “transforming” anything. And more concrete demonstrations of value.
Roy-Chowdhury will get more chances to make Grammarly’s case. The question is whether he — and the broader AI industry — will learn from this particular rake. The internet is watching. And it has Grammarly installed.
WebProNews is an iEntry Publication