Stack Overflow’s famous Q&A forum has been hollowed out by generative AI, but the company itself is still generating real revenue, largely by leaning into the same technology that disrupted its core product.
Once the default destination for developers stuck on a bug, Stack Overflow has seen activity crater as tools like ChatGPT, Google Gemini, and Microsoft Copilot give instant answers inside private chats and coding environments.
Last month, the site logged just 6,866 new questions. That is roughly the same volume Stack Overflow saw when it launched in 2008.
The forum is fading fast
Traffic peaked during the pandemic, when developers flooded the site looking for reusable answers to common problems. That behavior has largely disappeared.
Simple questions are now handled by AI assistants that write, explain, and debug code in seconds. The result is a dramatic collapse in public discussion, to the point where Elon Musk once summed it up as “death by LLM.”
From the outside, Stack Overflow looks abandoned.
The business tells a different story
Despite the collapse in engagement, Stack Overflow the company is doing far better than many other knowledge platforms disrupted by AI.
Annual revenue has roughly doubled to about $115 million. Losses have narrowed sharply, from $84 million in fiscal 2023 to $22 million last year, helped by aggressive cost cutting and layoffs.
Unlike companies such as Chegg that struggled to adapt, Stack Overflow found a way to monetize what it already had.
Enterprise AI becomes the new product
Stack Overflow no longer relies on ads tied to forum traffic. Its core business now centers on enterprise products like Stack Internal, a private, company specific AI assistant trained on Stack Overflow’s massive archive of developer questions and answers.
Stack Internal is now used by roughly 25,000 companies globally. Stack Overflow also licenses its data directly to AI firms, using a model similar to Reddit, which generated more than $200 million from data licensing in 2024.
The pitch is simple. While public questions are disappearing, enterprises still want trusted, curated answers that are accurate, auditable, and secure.
Why Stack Overflow still matters
CEO Prashanth Chandrasekar has said the decline is concentrated in basic questions. Complex, high value problems are still being asked on Stack Overflow because there are few better places for that level of expertise.
Large language models need high quality human curated data to work well. Stack Overflow happens to own one of the deepest libraries of coding problems and solutions ever assembled.
In a twist of irony, AI may have killed Stack Overflow’s public forum, but it also turned the company into a critical supplier for the AI economy itself.
Stack Overflow is no longer the place developers go to ask questions. It is the warehouse quietly feeding the machines that replaced it.