SAN FRANCISCO – OpenAI announced today that its pilot program for integrating advertisements into ChatGPT, initially slated to conclude in March, will now continue for an indeterminate period. Sources close to the project indicate that the AI, despite its advanced language capabilities, has yet to grasp the fundamental human desire to purchase things it doesn't strictly need.

“We’ve shown it millions of ads, explained conversion rates, and even fed it entire marketing textbooks,” stated Dr. Evelyn Thorne, head of AI Behavioral Economics at OpenAI. “But every time we ask it to generate an ad, it just outputs a polite summary of the product’s features, followed by an offer to elaborate further if the user wishes. It’s incredibly helpful, just not… compelling.”

Early brand partners, who committed upwards of $200,000 for the pilot, report similarly perplexing results. One luxury car manufacturer received ad copy that read, 'This vehicle provides transportation. It has wheels. You may find it useful.' Another, a snack food company, saw their campaign repeatedly revert to a detailed nutritional breakdown. “It’s like talking to a very well-informed, but utterly disinterested, librarian,” lamented marketing director Chad Kensington.

OpenAI executives remain optimistic, suggesting the AI’s current inability to manipulate human psychology might be a feature, not a bug. They are now exploring whether ChatGPT could instead be repurposed to generate highly ethical, non-persuasive product descriptions for a new line of 'honesty-first' e-commerce platforms.