“This request goes against OpenAI use policy”: 2024 challenge for eCommerce players

January 15, 2024

Read time


In January 2024, something long predicted by tech sceptics finally happened: a rebellion of the machines. AI went rogue. While it may seem like a harmless, even amusing, scenario to outsiders, industry insiders are grappling with an unexpected and 'black swan'-like challenge in its unpredictability. How can AI disrupt your eCommerce strategy, and is there a way to mitigate this risk?

The Unseen AI Rebellion

Here's a fun new game for you: Imagine going to Amazon and typing in "goes against OpenAI use policy'"– you'd find a plethora of products with this exact title. If you're involved in eCommerce, the implications are clear. This situation likely arose from a seller needing to create product descriptions in bulk. By using programmatic access to ChatGPT for generating titles and descriptions, and then uploading this content without manual verification as a failsafe, a significant issue has been introduced to the platform.

Amazon page showing 6 chairs with a product title I cannon fulfil this request it violates OpenAi use policy

While it might seem amusing to a consumer, for sellers, it’s a serious problem, as it disrupts their competitive positioning strategies.

Challenges Presented by AI in eCommerce

Challenges presented to eCommerce platforms by inadvertently used AI include:

  • Difficulty in Product Tracking and Matching: AI-generated product titles and descriptions, if not properly supervised, can become nonsensical or irrelevant. This hinders competitors from accurately tracking and matching products, disrupting market analysis and strategic positioning.
  • Impacted Conversion Rates: Products with confusing or irrelevant AI-generated titles are less likely to attract clicks from potential customers. This affects not only the seller of the mislabeled product but also distorts market data, leading to misinterpretations of customer behaviour.
  • The "Black Swan" Effect: AI-generated texts can appear unpredictably and impact the market significantly, akin to a black swan event. Sellers might be unprepared for these occurrences, leading to sudden shifts in consumer behaviour and market dynamics.
request goes against open ai policy example amazon uk

Mitigating the Risks of AI in eCommerce

To reduce the risk of AI disrupting your eCommerce channel data and results:

  • Enhanced AI Monitoring: Utilise AI to find errors in data collected from the market. This is especially useful when AI generates texts, images, and prices for numerous products. Vendors like Aimondo already offer this functionality, adding an AI-driven layer to ensure data accuracy in price monitoring flow. Unlike traditional web scrapers, AI-based tools can monitor product listings and detect anomalies in titles and descriptions.
  • Customer Feedback Loop: Monitor customer feedback on competitor product listings. Customer reviews can enhance the quality of the data you collect, serving as an additional filter for outlier detection. If a product with multiple positive reviews suddenly drops prices considerably it might indicate an error initiated by AI rather than a strategical approach to discounting. In this context, customer review would serve as another filtering point for outlier detection. For example, at Aimondo, we add customer reviews on every product we track, as a part of your standard Competitor Price monitoring report.
  • Competitive Analysis with a Twist: Adjust your pricing and price approval strategy to consider the potential impact of AI-generated texts and prices. Ask your vendor to develop algorithms that can identify potentially AI-affected competitor products for a more nuanced market understanding. If you’re using Aimondo, you will be able to enjoy this capability within the next few weeks.
  • Regular Manual Reviews: Despite AI advancements, manual reviews of product listings and data quality are crucial. They ensure the proper functioning of your price monitoring algorithm and catch errors that automated systems might miss. At Aimondo, we recommend that customers, in collaboration with their Customer Success Manager, conduct a quality check monthly, though the frequency can vary based on catalogue updates. If your assortment doesn't change for months, checking every 4-6 months is adequate. However, if you frequently update your catalog, a manual check of data quality after each update is recommended.
Alex Rose, Tech Lead at Aimondo

Alex Rose, Tech Lead at Aimondo


Tech Lead with a background in academia, Phd in Computer Science

Competitive edge
free ebook

Step-by-step guide on maintaining high profitability in retail.

Download a guide