Plattform
OpenAI
OpenAI ist das KI-Forschungs- und Bereitstellungsunternehmen hinter ChatGPT und definiert sich heute als Konversationsschicht, über die Käufer Produkte entdecken und Kaufabsichten äußern, während die eigentliche Transaktion an markenkontrollierte Systeme abgegeben wird. Seine Architekturentscheidungen verändern grundlegend, wie Einzelhändler über die Frage nachdenken müssen, ob sie den Checkout-Prozess selbst besitzen oder lediglich Distribution mieten.
Produkte & Dienstleistungen
ChatGPTmerchant checkout app integration (forthcoming)Instant CheckoutAgentic Commerce ProtocolOperatorComputer Using Agent (CUA)Checkout specification (discount and items_discount fields)Agentic Checkout SpecApps SDK UI guidelinesSupply Co. apparel line (T-shirts, socks, half-zip, basketball)Shopping with ChatGPT searchC2PA metadata and invisible watermarkingNegative Phrasescontext hintsad draft toolGPT 6 LunaGPT-6 SolGPT Image 1.5GPT Image 2o3GPT-5Ads Manager BetaOAI-SearchBotGPTBotgpt-4.1
Erwähnt in Rack & Reason (33)
- A Test Gate Read 800 Characters of Each Message. The Order Status Sat at Character 950.
- Muse Lost eBay. Its Robots File Now Lets ChatGPT and Claude Search.
- Apparel Leads a Google Ad Slot No Brand Can Buy or Measure Separately
- ASOS Credited Personalised Search With £31m. Its Try-On Got Adjectives.
- Otto Lent ChatGPT 19 Million Products. The Till Stays in Hamburg.
- Target's Title-Reading LLM Judge Missed Four in Ten Bad Recommendations on Its Benchmark
- Software Scores Midjourney's Faces as Less Idealised. People Rate Them Prettier.
- Seven Thousand Agent Sessions Ended In Seven Thousand Bookings
- Amazon Holds 63% of German Online Sales. The Search Box Is Where Rivals Could Cut In.
- ChatGPT's Picture Ads Fill the Wait for a Picture the Brand Never Sees
- ChatGPT Kept the Shopper's Face and Dyed the Skirt Indigo
- In Clothing, an AI-Visibility Score Measures the Questions
- AcCoRD Scored Four Kinds Of Shopper. Agents Failed The One Apparel Sells To.
- Nobody Abandons a Basket They Were Never Shown
- Nobody Ran the Red Team. The Counterparty Did.
- Ten Languages Went Into the Benchmark. Commerce Came Out Worst.
- The AI Labs Are Selling Hoodies. The Drop Mechanics Are Better Than Yours.
- The Lab Has No Store. It Has a $239 Chore Coat.
- Retailers Won Back the Checkout. The Agent Kept the Failures.
- The Agent Can Fill a Cart. It Cannot Give Up a Coupon.
- Cut the Dress for a Frame You Don't Own
- Agents Were Handed the Price. The Benchmark Asks If They Can Hold It.
- Your Promotions Were Priced for the Customer Who Forgets
- Ranking Was a Property Right. Inclusion Costs 4 Percent.
- The Agent's Checkout Belongs to the Company That Files the Duty
- The Agent Issued the Refund. Only a Gate Caught the Breach.
- Doubao and Qwen Are Deleting Their Users' Agents. Retail Budgeted the Cost at Zero.
- The Shopping Agent Has to See Your Screen. The Retailer Owns What It Sees.
- John Lewis Funded AI Shopping and TikTok Shop From the Same £800m
- John Lewis Paid Retail For The Funnel £800m Couldn't Build
- Sephora Saw What Instant Checkout Missed
- If ChatGPT Gets the Product Wrong, Fashion Still Gets Blamed
- OpenAI Blinked on Checkout — Fashion Brands Now Have to Decide What That Means for Them
Erstmals erwähnt: März 2026