OpenAI启动新广告格式，在ChatGPT图像生成结果旁展示赞助商产品图片，初期仅限美国。
OpenAI has begun testing a new visual ad format in the US: sponsored product images displayed adjacent to DALL·E-generated outputs—triggered contextually (e.g., showing a coffee maker after generating a kitchen scene), not via keyword matching. This represents the first large-scale deployment of native visual advertising within a mainstream AI chat interface. For early-stage AI teams, this signals two strategic shifts: First, monetization is pivoting decisively toward low-friction, flow-native models—ads embedded directly in existing UX flows (no redirects, no new screens) dramatically reduce CAC and increase yield per active session. Second, multimodal understanding + generation capabilities are becoming direct revenue levers: teams that own the full image pipeline can define custom ad triggers, negotiate premium CPMs with vertical brands, and retain full user intent data. We advise startups to architect ad insertion points from v1—not as an afterthought—and prioritize high-intent use cases (e.g., home renovation, medical imaging prep) where visual relevance drives measurable conversion lift.