The launch builds on the company’s recent product momentum, including last month’s debut of ChatGPT Atlas, an artificial intelligence (AI) browser designed to turn the web into a working environment rather than a collection of tabs.
OpenAI is moving to embed its ChatGPT platform more deeply into professional and personal workflows with the pilot launch of a new group chat feature. The testing leverages AI’s power to coordinate, plan, and execute group projects in real-time, directly targeting the organizational and communication needs of professionals across the banking, payments and FinTech sectors.
The new feature allows users to bring up to 20 people into a shared digital space to collaborate with each other and with ChatGPT simultaneously. Users can initiate a group chat by tapping a new people icon, which creates a copy of an existing conversation or a new one, ensuring the user’s personal chat history and memory remain private and unshared with the group. OpenAI is rolling out this pilot to users on ChatGPT Free, Go, Plus, and Pro plans in Japan, New Zealand, South Korea and Taiwan for now. The more universal GPT-5.1 release this week unveiled two versions for users.
The New GPT-5.1
GPT-5.1 arrives in Instant and Thinking. Instant is now the default model across paid ChatGPT tiers and is built for everyday speed and conversational use. Thinking supports more complex reasoning and adjusts its computation time based on task difficulty. OpenAI said Instant delivers a “warmer” tone and stronger adherence to instructions while maintaining a faster response rate.
The update also expands ChatGPT’s personalization features. Users can choose new tone presets, including Friendly, Professional, Candid and Nerdy, with experimental controls available for warmth and conciseness. The customization tools follow the design approach seen in ChatGPT Atlas, which integrates AI directly into browsing to create a single workflow that can summarize pages, extract insights or complete actions without switching contexts.
Advertisement: Scroll to Continue
The GPT-5.1 release comes as OpenAI reports surpassing 1 million paying business customers. The company said organizations across life sciences, retail, technology and financial services sectors are adopting its models for research assistance, summarization and customer operations. The milestone signals that enterprise AI deployment is moving deeper into production environments, supported by higher-tier features and more predictable model behavior.
The competitive backdrop is shifting as well. PYMNTS reported that Anthropic is on track to reach profitability sooner than OpenAI, highlighting the economic pressure behind the industry’s rapid expansion. Anthropic expects to break even around 2028, while OpenAI’s forecast shows substantial operating losses through the decade. The contrast underscores two diverging strategies: OpenAI is prioritizing scale and product breadth, while Anthropic is focusing on cost efficiency and a narrower enterprise footprint.
Taken together, GPT-5.1 reflects a turn toward usability rather than pure benchmark gains. Its dual-track design expanded tone controls and improved alignment indicate a system optimized for reliability in everyday workflows. For enterprises evaluating long-term AI investments, the release adds to OpenAI’s growing platform but also arrives at a moment when cost discipline and vendor economics are becoming central to adoption decisions.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
