Testing ads in ChatGPT
| Source: Mastodon | Original article
OpenAI has begun a live test of advertising within the ChatGPT interface, marking the company’s first overt move to monetize the free tier beyond its subscription‑only “ChatGPT Plus” model. In a brief blog post the firm described the experiment as an “ads pilot” aimed at “supporting broader access to ChatGPT while preserving consumer trust, usefulness, and user control.” Early internal metrics, according to the statement, show “no impact on conversation quality” and a “positive response” from the limited user group that has been exposed to the new slots.
The rollout arrives at a moment when OpenAI’s operating costs are soaring. Training the latest GPT‑5.4 model, announced last week, reportedly cut hallucinations by 30 % but also required a larger compute budget. Advertising revenue could help offset those expenses and keep the free tier viable, a goal that aligns with the company’s public pledge to democratise AI. Yet the move also fuels long‑standing worries about “enshittification” – the gradual degradation of a platform as it prioritises profit over user experience. Critics argue that even well‑intentioned ads risk cluttering the conversational flow, nudging users toward sponsored content, and creating new privacy vectors as OpenAI gathers data to target those placements.
What to watch next includes the geographic scope of the pilot and the formats being tested, from banner placements to in‑response suggestions. User sentiment will be gauged through feedback loops and churn rates on the free tier, while regulators in the EU and Norway may scrutinise how ad‑driven data collection complies with emerging AI‑specific privacy rules. Competitors such as Google Gemini and Microsoft Copilot are likely to monitor OpenAI’s approach closely, as any shift in the economics of conversational AI could reshape pricing and feature strategies across the sector.
Sources
Back to AIPULSEN