OpenAI’s latest launch promises to offload entire workdays onto AI agents, even as office workers and IT leaders scramble to understand which new model, mode, or “work loop” they’re supposed to trust.

Late June–Early July: From Regulatory Drama to Public Rollout

After a brief period in which GPT‑5.6 was available only to government‑approved organizations, the Trump administration cleared the model for public release, prompting OpenAI to call it “the best model we have ever produced.”¹ The company simultaneously unveiled ChatGPT Work, an agent that “can gather context from the apps, files, and workflows you choose and create finished materials such as documents, spreadsheets, presentations, and web apps.”¹

OpenAI’s own blog framed GPT‑5.6 as its flagship series, designed to deliver “more useful work from every token, with stronger performance per dollar and on demand capability for the most complex tasks.”²

July 9: GPT‑5.6, ChatGPT Work, and the Office Land Grab

On July 9, OpenAI formally launched the GPT‑5.6 family—Sol, Terra, and Luna—positioning the variants for enterprise work, coding, and scientific research, and calling 5.6 its “strongest cybersecurity model yet, achieving frontier performance with significantly fewer tokens.”³

Business press described the release as “a super app for work,” as OpenAI merged its Codex coding tool into the ChatGPT desktop app and revealed a new Work setting, part of an “ambitious strategy” to become “an indispensable part of life for millions of workers.”⁴ Ars Technica stressed that ChatGPT Work is pitched as a tool that can “stay with a project for hours if needed, and turn a goal into finished work,” including Scheduled Tasks that “take repetitive tasks off your plate.”⁵

At the same time, OpenAI introduced GPT‑Live, “a new generation of voice models for natural human‑AI interaction, now powering ChatGPT Voice.”⁶

Enterprise Perspective: Microsoft, Amazon, and Perplexity Move Fast

OpenAI and Microsoft sought to quash speculation about a rift when OpenAI announced that GPT‑5.6 “will become the new preferred model in Microsoft 365 Copilot—in Word, Excel, PowerPoint, Chat and Cowork,” promising higher‑quality documents, analyses, and presentations “with less effort” for Microsoft 365 users.² TechCrunch noted that this “preferred model” status reassured observers that OpenAI’s software would continue to power Microsoft’s workplace tools, even as Microsoft develops its own MAI models.⁷

Amazon quickly followed: AWS touted that “GPT‑5.6 Sol, Terra, and Luna from @OpenAI are now generally available on Amazon Bedrock,” offering “three tiers of intelligence, from flagship reasoning to fast inference.”Greg Brockman (gdb)⁸ Perplexity highlighted that GPT‑5.6 “sits on the Pareto frontier of our agentic research evals: higher accuracy at lower cost,” and updated its Agent API accordingly.Aravind Srinivas (AravSrinivas)⁹

Human Workers: From Task Doers to System Gardeners

Commentators on knowledge work argued that GPT‑5.6 “is the first model I’ve used that can reliably run whole loops of knowledge work, not just help with individual tasks,” shifting jobs “from doing the work to tending the system that does it.”¹⁰ One piece framed this as a move toward “the knowledge work loop, with human judgment at the center,” where Sol monitors inboxes, drafts replies, and learns from user feedback over time.¹⁰

Yet this power comes with complexity. Axios described “ChatGPT’s paradox of choice,” as workers face a “jumble of new model names, reasoning levels and subscription tiers — and [are] left… to figure out which combination fits their work and budget.”¹¹

Reviewers gave Sol strong marks for collaboration: one outlet’s “vibe check” concluded that “GPT‑5.6 Sol Is Our Favorite Model to Collaborate With,” while still preferring rival Claude Fable for fully delegated, hands‑off tasks.¹²

AI’s Own Pitch: Agents at Consumer Scale

OpenAI’s product narrative emphasizes that ChatGPT Work “can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work,” powered by GPT‑5.6’s ability to reason through “multi‑step tasks and create materials that follow your templates and reference files.”¹³ The company stresses that users can “follow its progress, answer questions, change direction, and approve important actions,” framing the agent as a controllable partner rather than an unchecked automation layer.¹³

Executives describe this as bringing “agents to consumer scale,” allowing people to run serious workflows “from your phone, no laptop required.”Greg Brockman (gdb)¹⁴

A Crowded, Accelerating Race

Business Insider captured the broader context as part of a “whirlwind 72 hours of rival AI announcements,” with OpenAI’s GPT‑5.6 and ChatGPT Work going up against Meta’s Muse updates, SpaceXAI’s Grok 4.5, and Anthropic’s new features.¹⁵ Another report concluded that OpenAI is “making its biggest play for the office,” as it fights to “stay at the forefront of AI capabilities, duking it out release by release with rival Anthropic” and others.⁴

For now, GPT‑5.6 and ChatGPT Work offer a glimpse of offices where much of the doing is automated—but where humans still have to decide which AI to hire, and how tightly to hold the reins.