Key Takeaways
- According to OpenAI’s new “GDPval” benchmark, its latest models can compete with experienced professionals across 44 different jobs, from lawyers to software developers.
- A new “Instant Checkout” feature allows users to buy products from Etsy and Shopify merchants directly within a ChatGPT conversation.
- In OpenAI’s own test, rival Anthropic’s Claude model actually scored higher, highlighting the intense competition in the AI space.
OpenAI made two major announcements this week, claiming its ChatGPT models can now rival human experts across dozens of professions while also launching a new feature that lets users shop and check out directly within the chat interface.
The company’s bold performance claims come from a new benchmark called GDPval, designed to measure AI’s ability to perform real-world work tasks. The framework tested AI on 1,320 assignments across nine major industries, with human professionals averaging 14 years of experience evaluating the results. While OpenAI’s GPT-5 showed significant improvement, rival Anthropic’s Claude Opus 4.1 actually outperformed it, achieving a higher win-and-tie rate against the human experts.
on economically valuable tasks compared to an industry professional.
OpenAI says its models can complete these professional tasks 100 times faster and cheaper than their human counterparts, marking a significant step in the company’s quest to measure AI’s real-world economic impact.
On the commerce front, OpenAI rolled out “Instant Checkout,” a new feature that integrates e-commerce directly into ChatGPT. U.S. users can now find and purchase items from Etsy sellers using natural language prompts like “find a gift for a ceramics lover.” Over one million Shopify merchants, including major brands like Skims and Glossier, are set to join the platform soon. The move, which caused a significant jump in stock prices for both Etsy and Shopify, is powered by a new protocol developed with Stripe.
These developments arrive as the tech industry grapples with proving the return on investment for AI. With competitors like Anthropic pushing the boundaries of autonomous AI and Microsoft building its own in-chat commerce tools, OpenAI’s latest moves signal a clear strategy to embed its technology deeper into both professional workflows and consumer life. However, OpenAI acknowledges the current limitations, noting that real-world jobs are far more complex than a collection of well-defined tasks.
Your First 10 AI Skills
10 practical AI skills, copy-paste prompts and a 7-day plan to start using AI with confidence.
Download the guide →