AI Daily · Nov 14, 2026: ChatGPT Wallet opens agent payments to US Pro users
ChatGPT Wallet rolls out to US Pro subscribers, letting agents complete purchases with approval and no card exposure; Command Code's $10 plan raises GLM-5.3 credits from $20 to $70.
Pick a date
-
ChatGPT Wallet lets agents complete purchases for Pro users in US
ChatGPT Wallet is rolling out to Pro subscribers in the US, letting agents complete purchases with user approval; card details are never exposed and every purchase requires a security check. It is being built with American Express, Mastercard and Visa.
-
Command Code GOAT $10 plan raises GLM-5.3 credits to $70
Command Code's $10/mo GOAT plan now includes $70 of GLM-5.3 usage (up from $20), $100 of GLM-5.3 Flash and $100 of DeepSeek V4.1 Flash 1017, plus 72 other models, with 1M token context and more for Pro and Max.
-
Cursor publishes update on recent security incident
Cursor released a detailed overview of a recent security incident and the remediation steps it has taken.
-
Perplexity adds Teammates agents to Mac app for Max subscribers
Perplexity's Teammates feature is now available in the Mac app for Max subscribers; the agents can work alone or together on QA, customer service, research and growth marketing tasks.
-
Hugging Face tool exports, quantizes and runs any transformers model on any backend
A new Hugging Face release lets users export, quantize and run every transformers model end-to-end on any backend with identical inputs and outputs as a drop-in replacement.
-
TRL v1.18.0 stabilizes AsyncGRPOTrainer with parallel generation and training
AsyncGRPOTrainer is now stable in TRL v1.18.0, running generation and training simultaneously on separate GPUs rather than taking turns, with documentation on usage and comparison to GRPOTrainer.
-
LiteLLM launches private registry with early severe vulnerability fixes
LiteLLM Private Registry is available to customers holding an existing LiteLLM confidentiality agreement, offering early fixes for the most severe vulnerabilities while all security fixes still go public.
-
aimlapi runs SEO campaign for $0.61 using Jev keyword scoring
aimlapi says an agent can analyze a site and competitors' ads, score over 10k keywords with TypeSafe AI's Jev, and launch a Google Ads campaign for $0.61 with 0% markup, versus $349/mo tools.
-
Anthropic publishes 10 recommendations for designing and reporting safety evaluations
Anthropic's Engineering Blog outlines 10 recommendations for designing, running and reporting safety evaluations that judge whether a model is ready to deploy.