ChatGPT ad vs Claude Ad test
On February 9, 2026, OpenAI quietly began testing ads inside ChatGPT. They showed up for logged-in users on the free and lower-cost U.S. tiers, and stayed away from sensitive topics like health, politics and finance. What really set people off was a question of trust and privacy: the worry that a conversation you treat as private might quietly shape what the model shows you. Anthropic answered within days. It positioned Claude as ad-free and argued that ads have no place in an open-ended conversation, a line the press read as a direct shot at OpenAI. Within a week, the two leading AI labs had picked opposite sides: monetised reach for one, ad-free trust for the other.
To see how that split actually landed, Pulsar tracked more than 130,000 conversations across 16 weeks of weekly net sentiment, using Pulsar TRAC and SAGA, our insight agent. The reception was lopsided. Claude ran net-positive in 15 of 16 weeks, while ChatGPT spent the first half of the window underwater. The gap peaked at +59.9pp on March 22 and flipped just once, the week Anthropic shipped its own "Claude Ads" tool. The two brands also drew very different critics. ChatGPT took a privacy and surveillance backlash, while Claude took a military and geopolitical one. Underneath the split, the same shift was unfolding for both: the early skepticism about AI putting ads into the conversation was starting to look like opportunity, especially for the brands and strategists paying attention.
Weekly net sentiment · Feb to May 2026
Sentiment toward ChatGPT and Claude ads is diverging.
Weekly net sentiment (positive minus negative, in percentage points) for each brand’s advertising narrative. Claude ran net-positive for fifteen of sixteen weeks. ChatGPT spent the first half of the window underwater.
Net sentiment in percentage points. Annotations mark the three turning points that defined the window. Zero line = balanced sentiment.
View the weekly net-sentiment data (percentage points, 16 weeks)
| Week (2026) | ChatGPT / OpenAI | Claude / Anthropic |
|---|---|---|
| Feb 1 | +4.3 | +10.2 |
| Feb 8 | -32.7 | +6.0 |
| Feb 15 | -18.5 | +21.9 |
| Feb 22 | -10.3 | +14.7 |
| Mar 1 | +6.4 | -10.9 |
| Mar 8 | -22.4 | +22.7 |
| Mar 15 | -13.2 | +30.9 |
| Mar 22 | -10.3 | +49.6 |
| Mar 29 | +0.9 | +35.2 |
| Apr 5 | +0.5 | +36.4 |
| Apr 12 | +4.2 | +38.2 |
| Apr 19 | +16.6 | +56.3 |
| Apr 26 | +56.2 | +67.3 |
| May 3 | +28.4 | +63.7 |
| May 10 | +23.9 | +62.3 |
| May 17 | +35.1 | +43.7 |
Look at ChatGPT on its own and the story gets even clearer. From the pilot’s February 9 launch, ad sentiment climbed from −17.9% to 49.3% by mid-April, a 67.2-point swing from backlash to acceptance, as marketing and business communities became the loudest positive voices. Pulsar’s data on that shift was featured in Digiday.
“As the conversation expanded beyond core AI users, marketing and business communities became some of the strongest positive voices, reframing ads as a sign of platform maturity rather than decline.”Dahye Lee, Senior Marketing Innovation Lead, Pulsar
Part of that lift is simply the noise settling. Even as the conversation broadened to marketers and business voices, the sheer volume around the ads has come well off its February peak, and as the early backlash quieted, sentiment had more room to climb. A calmer read does not mean the product is sorted, though. OpenAI’s pilot is still working through under-delivery, limited inventory and early-stage infrastructure, with the company prioritising caution over scale. That gap between perception and execution is still the real tension in the rollout.
Drivers by phase · Feb to May 2026
What actually moved the sentiment.
Two troughs and one shared peak account for most of the swing. These were the moments that carried the whole window.
● ChatGPT: deepest dip
● Claude: only negative week
● ChatGPT: best week
● Claude: peak
ChatGPT / OpenAI · top voices
Who’s driving the ChatGPT story?
Ranked by visibility (followers × reach × amplification). Positive voices skew toward spectacle and OpenAI’s own comms. Negative voices cluster around one frame: surveillance.
The catalyst · negative
Claude / Anthropic · top voices
Who’s driving the Claude story?
A completely different cast. Positive voices split into a Super Bowl brand wave (Feb) and a practitioner-builder wave (Apr to May). Negative voices drop privacy entirely and go after Anthropic on military and geopolitical grounds.
The brand line · positive
Strategic read · May 2026
So what does it all mean?
Sentiment is warming for both brands, but each carries a very different risk underneath. Here is what could pull each one back down.
To stay up to date with our latest insights and releases, sign up to our newsletter below:
