By opting for a Cost Per Mille (CPM) model—where advertisers pay for every 1,000 views—OpenAI is signaling that the value of an AI ad lies in its contextual relevance rather than immediate traffic diversion.
The Logic Behind the CPM Model
- Lower Click Intent: Internal research suggests that AI users are in a “research and decision” mode, making them less likely to click away to a third-party site compared to traditional search engine users.
- Conversational Immersion: OpenAI wants ads to feel like helpful suggestions at the bottom of a response. A click-based model might incentivize “clickbait” formatting, which would degrade the user experience.
- Future Interactivity: OpenAI plans to eventually make ads interactive, where users can ask follow-up questions directly to the sponsored unit (e.g., “Does this discounted hotel have a gym?”) without ever clicking a link.
Ad Rollout: Tiers and Targets
The ad testing phase, beginning in February 2026, is strictly controlled to prevent user backlash.
| Feature | Details |
| Affected Tiers | Free Tier and the new $8/month “ChatGPT Go” plan. |
| Ad-Free Tiers | Plus ($20/mo), Pro ($200/mo), Team, and Enterprise. |
| Placement | Separated box at the bottom of the answer; clearly labeled “Sponsored.” |
| Targeting | Contextual only (based on current chat); no behavioral tracking. |
The “Instagram” Inspiration
CEO Sam Altman has cited Instagram as his inspiration for the ad model, stating that he wants ads to help people “discover stuff they like that they otherwise never would have,” rather than functioning as disruptive banners.
The Financial Necessity: $14B in Losses
The pivot to an ad-supported model is a direct response to OpenAI’s soaring infrastructure costs.
- The $14B Gap: Despite hitting a $20 billion revenue run rate in 2025, OpenAI is projected to lose $14 billion in 2026 due to massive GPU and energy requirements.
- The 2029 Goal: OpenAI expects its advertising and e-commerce business to contribute over $25 billion to its annual revenue by 2029.
Guardrails and Restrictions
To maintain “Answer Independence,” OpenAI has implemented strict safety protocols for 2026:
- No Under-18s: Ads will not be shown to users predicted to be minors.
- Sensitive Exclusions: Ads are strictly prohibited in conversations regarding health, mental health, or politics.
- Integrity Guarantee: Advertisers cannot pay to influence the actual text or “opinion” of the AI’s response.
Conclusion: Redefining Digital Marketing
By charging per view, OpenAI is challenging the dominance of Google’s “Cost Per Click” (CPC) philosophy. If successful, this CPM approach could turn ChatGPT into a high-premium “brand awareness” engine where companies pay to be the recommended solution in the world’s most popular AI interface.
