AI Morning Brief: OpenAI Blocks Ads in Image Products for Adobe and Others; Qwen’s New AI Glasses Leak With Iris Payment

Avatar 0

OpenAI Blocks Adobe and Others From Advertising in Image Products

On September 9 local time, NUPIAO learned that Doug Wyatt, Adobe’s senior director of media for the Americas, said OpenAI has told some commercial partners it will no longer accept ads in ChatGPT’s image and audio generation products. Word is Adobe took part in a ChatGPT ad pilot earlier this year.

Qwen’s New AI Glasses Spotted for the First Time, Support Iris Payment

On September 10, NUPIAO spotted what looks like Qwen’s new AI glasses at the AI payment zone of the 2026 Bund Conference. Our team on the ground saw the glasses largely wrapped in black concealing material, so we couldn’t make out more features or hardware details just yet. But based on what’s in the exhibit, the new product may bring in iris recognition for the first time, using more precise identity matching to boost both security and convenience for AI glasses payments. 

Moonshot AI Rumored to Plan Dual Listing in Hong Kong and Shanghai; Source Says It’s Not True

On September 10, NUPIAO heard reports that Moonshot AI, a top Chinese large-model player, is weighing a dual listing in Hong Kong and Shanghai to raise more cash. A person familiar with the matter, however, says the rumor is not accurate. 

Alipay Launches “Agent Emergence Award” With 10 Million Yuan Annual Fund

On September 10, Alipay announced the official launch of its “Agent Emergence Award,” pouring 10 million yuan a year into a development fund, with a 1 million yuan annual grand prize. The award runs in two stages: quarterly shortlists and an annual contest. First-batch shortlisted works get a 5,000 yuan milestone reward, and the annual grand prize of 1 million yuan is expected to be announced in Q4 2027.

Anthropic Opens Mythos Model Access to the EU

On September 10, NUPIAO learned that Anthropic has opened access to its Mythos AI model for the EU’s cybersecurity agency.

European Commission spokesperson Thomas Regnier said: “After constructive talks with Anthropic, we can confirm that the European Union Agency for Cybersecurity (ENISA) has been granted access to Mythos 5 and is currently testing it.” Regnier said ENISA won’t get access to the model’s latest version, Mythos 5.1.

DeepSeek V4.1 Flash Officially Released

On September 10, DeepSeek V4.1 Flash was officially released. According to the company, the model features native multimodal visual understanding. The new architecture was designed for a higher capability ceiling, faster inference, greater throughput, and scalability to larger parameter models. Thanks to the architectural innovation, DeepSeek has lowered V4.1 Flash pricing. And to allocate resources more reasonably, it still uses peak and off-peak pricing, with off-peak rates at half the peak price, encouraging users to adjust task timing based on actual usage. The new pricing takes effect at 12:00 on September 10, 2026.

Advertisement

Leave a Reply

Your email address will not be published. Required fields are marked *

Log In / Sign Up

Enter code for secure login, or use password.

Code Login Password Login