Google AI Mode shows same products 21.6% more expensive than search
Google's AI Mode might be making shopping more expensive for users. A new data study by Productrise tracked over 2 million product listings across more than 100,000 SERPs and AI Mode responses over 23 days (August 9–31, 2026). When the exact same product appeared in both AI Mode and traditional search, the price in AI Mode was on average 21.6% higher.
Key Findings
- For matched products (appearing in both), AI Mode prices are 21.6% higher.
- Looking at all listings, AI Mode's median price is $149 vs $100 in traditional search — about 49% higher.
- Only 1.28% of products ranking in traditional search also appear in AI Mode for the same query on the same day.
- Price discrepancies occur 38.1% of the time on matched products; when they do, AI Mode is more expensive 68.4% of the time.
- The main seller differs on 49.6% of matched products.
The Data
The study ran the same shopping queries through both Google AI Mode and traditional search at the same moment. For price comparisons, they matched identical products. The 21.6% price gap on matched items shows a clear pattern: when the algorithm has a choice, it tends to surface the pricier listing.
This effect is even stronger when looking at the overall mix of products. AI Mode shows far fewer products — on average 3.9 vs 27.8 in traditional search — and those products skew toward higher price points. The median product in AI Mode is nearly 50% more expensive than its traditional counterpart. This suggests that Google's AI Mode is not simply returning the same results in a new format; it's curating a different, more expensive set of recommendations.
Why It Matters for Developers
For developers building shopping tools or price comparison sites, this study highlights a potential bias in AI-generated responses. If you're relying on AI Mode as a data source for price aggregation, you might be passing on higher costs to your users. Independent SEO consultant Brodie Clark notes that while the higher price might be displayed in the grid result, "the retailer with the lowest price gets the click in the end" — but that's only if the user sees the option. The study suggests that AI Mode curation could be steering users toward more expensive inventory without them realizing cheaper alternatives exist just a tab away.
This isn't just a consumer issue — it's a data integrity issue. If you're building on Google's AI Mode output, be aware that its "recommendations" may not be price-optimized. The study's authors argue that "open market comparison is replaced by a curated path to higher-priced stock." The full data distribution is available in the source below.
📖 Read the full source: HN AI Agents
👀 See Also

Inference Pricing Analysis Shows 4.4x Spread for Same Model Across Providers
Analysis of inference pricing for Llama 3.1 70B Instruct shows a 4.4x cost difference between providers, with DeepInfra at $0.20/$0.27 per million tokens and Together at $0.88/$0.88. For reasoning models, the spread reaches ~30x between DeepSeek R1 and OpenAI o1.

Claude Platform on AWS Now GA: Native Anthropic Experience via IAM, CloudTrail, and AWS Billing
AWS announced GA of Claude Platform on AWS, giving developers direct access to Anthropic's native Claude experience through existing AWS accounts with IAM auth, AWS billing, and CloudTrail logging — but customer data is processed outside AWS security boundary.

19 Days to a Reddit App: Prestige Tracks Community Contribution Beyond Karma
A developer built Prestige, a Reddit app for seasonal prestige, community rankings, and Hall of Fame tracking, using OpenClaw AI.

Claude Opus 4.7 Released with Hybrid Reasoning and 1M Context Window
Anthropic released Claude Opus 4.7, a hybrid reasoning model with a 1M context window that delivers stronger performance on coding, vision, and complex multi-step tasks. Pricing starts at $5 per million input tokens and $25 per million output tokens.