Google's deal with StackOverflow is the latest proof that AI giants will pay for data


Last year StackOverflow became one of the first websites to announce that it would charge AI giants for access to content used to train chatbots. Now the popular Q&A service for coders has signed on its first client—Google—which CEO Prashant Chandrasekhar says marks the beginning of a “meaningful” new stream of revenue.

The deal is significant because it is unclear how much Google and other AI developers will pay for the content needed for AI projects. Millions of books and websites have promoted the development of AI systems, but most publishers are not compensated, and some are suing over abuse. Many publishers, including StackOverflow, appear to be under threat from ChatGPIT and other generative AI products, which can answer questions that previously sent coders their way.

The deal will see Google's cloud division use Stack Overflow's questions and answers about Google Cloud services to provide coding help and technical support through a version of Google's Gemini chatbot. Google's cloud computing customers will also be able to ask questions through Google Cloud's command-line interface. “Their AI may not have all the answers, and so we have huge potential to help them close that loop,” Chandrasekhar says. “We are the largest place where community knowledge is compiled and validated.”

Gemini will summarize answers received from StackOverflow in its own words but include the company logo, a link to the original content, and the username of the site contributor who supplied it. The companies plan to demonstrate the system at Google Cloud Next, the search company's annual cloud conference, in April and launch it shortly thereafter.

Chandrasekhar says there are no significant restrictions on how Google can use Cloud Stack Overflow data, which means it can be used to train large language models and other AI systems. “Where we want to stand firm is the things that are non-negotiable for us – trust, accuracy, quality and attribution to the sources of these AI outputs,” he says.

He declined to say how much Google was paying StackOverflow for the data. “This will be a meaningful business offering for us in the near term, medium term and long term,” says Chandrashekhar.

covert scraping

Google and other AI developers have previously collected data from StackOverflow and other websites without notice. As demand for generic AI technologies has grown – and the valuations of the companies developing them have soared – websites that supply basic text have begun to demand what they see as their fair share. Chandrashekhar says, fortunately for StackOverflow, potential customers have heeded the message. “We don't have to chase people down,” he says.

StackOverflow data is particularly beneficial for AI systems that generate computer code, which has proven popular among software engineers and is an important source of revenue for Microsoft and OpenAI.

The new StackOverflow deal comes just a week after Google reached a licensing agreement to collect data from discussion forum operator Reddit, the content of which has helped chatbots' ability to interact. Reddit unveiled plans to start charging for data access just before StackOverflow last year.