Powered by the GPT-5.6 model family, ChatGPT Work combines connected workplace context, research, content creation, and task execution as OpenAI expands AI beyond conversational assistance
Dubai — OpenAI has outlined its expanding vision for agentic AI in the workplace, positioning ChatGPT Work as a system capable of moving beyond answering questions to researching, creating and completing complex professional tasks.
During a recent media education session, OpenAI demonstrated how ChatGPT Work brings together its latest GPT-5.6 models with connected workplace applications, files, research tools and automation capabilities.
The session was hosted by Andy Rogerson, EMEA Communications at OpenAI, and Conor Spicer, Technical Lead MENA at OpenAI, and explored the company’s latest developments across ChatGPT Work, the GPT-5.6 model family, and its broader portfolio of AI tools.
ChatGPT Work is designed as an AI partner for longer and more complex projects, gathering context from connected applications, files, tools and workflows before planning and executing work.
According to OpenAI, ChatGPT Work can create editable documents, spreadsheets, presentations, reports, charts and interactive Sites. Users can remain involved throughout the process, answering questions, redirecting the approach, reviewing outputs and approving important actions.
OpenAI said ChatGPT Work now has access to around 1,400 plugins, tools and skills that can bring an organisation’s existing information and workflows into the AI environment. Scheduled Tasks can also run work once, repeatedly according to a schedule or in response to a trigger, as well as monitor systems for changes. The company said more than 10 million people are now weekly active users of its agentic products.
GPT-5.6 expands model choices
ChatGPT Work is powered by OpenAI’s GPT-5.6 family, which includes Sol, Terra and Luna and is designed to offer different balances of intelligence, speed and cost.
GPT-5.6 Sol is positioned as the flagship model for advanced reasoning and complex professional workflows, with OpenAI highlighting its capabilities across coding, scientific research, data security analysis, and knowledge work. OpenAI reported that GPT-5.6 Sol achieved state-of-the-art results on DeepSWE v1.1, a benchmark for long-horizon software engineering tasks, and on the Artificial Analysis Coding Agent Index. On DeepSWE v1.1, Sol scored 72.7%, surpassing Claude Fable 5’s 69.9% at an estimated 36.2% lower API cost. On the Artificial Analysis Coding Agent Index, Sol achieved the highest score while using less than half the output tokens, taking less than half the time, and costing about one-third less than Fable 5.
GPT-5.6 Terra provides a more balanced, lower-cost option for everyday workloads, while GPT-5.6 Luna is positioned as the fastest and most cost-efficient model within the family.
According to OpenAI’s “How agents are transforming work” report, organizations are increasingly deploying AI agents to support work across functions including software engineering, customer support, finance, sales, and operations. The report highlights how AI is being used to automate repetitive tasks, assist with research and analysis, streamline workflows, and enable employees to focus on higher-value work.
Connecting AI to everyday work
A central theme of ChatGPT Work is connecting AI with the applications and information employees already use. Plugins can provide access to authorised information from workplace services, while users can also upload relevant files and provide project-specific context. ChatGPT Work can then search across these sources, analyse information and use it to complete a broader task.
OpenAI demonstrated use cases ranging from turning research and Slack discussions into presentations to searching workplace conversations for earlier decisions, creating live account hubs, generating meeting follow-ups and handling workflows across desktop applications and local files.
For journalists, OpenAI proposed several practical workflows. One example involves building a morning reporting brief using information from Gmail, Slack, Google Calendar and Drive alongside current news developments. The system could organise results around breaking developments, announcements under embargo, source messages requiring replies and upcoming meetings.
Another workflow could analyse calendar and email commitments to identify uninterrupted writing periods, while retaining human approval before changing or cancelling appointments. ChatGPT Work could also use a journalist’s articles, biography, CV, headshot and social links to assemble a personal portfolio website and present it for review before publication.
OpenAI’s starter guidance also highlights workflows including inbox management, meeting preparation, expense filing, converting numerical data into recommendations and transforming meeting notes into presentations.
Broader AI ecosystem
The session also highlighted OpenAI’s wider product portfolio, including Deep Research, ChatGPT Images, Codex, Study Mode, ChatGPT Edu, and Apps in ChatGPT.
Deep Research is designed to conduct multi-step investigations across webpages, documents, images and PDFs before producing structured, sourced reports. Codex extends agentic capabilities into software development and related workflows, supporting over 6 million weekly active Code users.
OpenAI said its products and platforms are supported by a global team of over 4,000 employees, including more than 650 AI researchers and over 11 global offices.
The company’s enterprise push is increasingly focused on shifting AI from an isolated productivity assistant towards systems capable of working across organisational information, applications and workflows while keeping users involved in reviewing and approving consequential actions.
Source: Tahawul Tech