Japan's CyberAgent - the company behind streaming service AbemaTV and one of the country's largest ad tech businesses - has published a case study on deploying ChatGPT Enterprise and OpenAI's Codex across teams in advertising, media, and gaming.
The two tools serve different purposes. ChatGPT Enterprise is a company-managed version of ChatGPT where employee conversations don't feed OpenAI's training data - which matters for any business handling client campaigns or proprietary content. Codex handles code generation, letting developers write and iterate faster than they could manually.
CyberAgent's stated goals were practical: scale AI adoption internally, improve output quality, and speed up decision-making. The deployment covers three business units with genuinely different day-to-day work - ad teams managing campaigns and creative, media staff producing content, game developers writing code.
Rolling out AI tools across diverse teams is harder than it looks. Most large companies end up with a patchwork of individual subscriptions, employees using personal ChatGPT accounts on company data, no visibility into what's happening. Standardizing on enterprise tools gives IT at least basic data controls and a record of how AI is actually being used.
OpenAI publishes case studies like this one as marketing material, so they're worth reading critically. But documented examples of broad internal AI deployment are useful for other companies still weighing whether the organizational overhead is worth it. CyberAgent's experience suggests yes - the question is which tools and governance model you pick.