Skip to content
All tool news

Tool desk · Updated daily

Tool news·ChatGPT·

ChatGPT Expands Codex Desktop App With Mac App Control and Task Automation.

ChatGPT's Codex desktop app now controls Mac applications, connects to external tools, and automates repeatable tasks with memory.

CW

Create With tool desk

1 source checked · 2 min read · 2 sections

ShareLinkedIn

OpenAI has begun rolling out a major expansion of capabilities for its Codex desktop app, transforming the AI assistant into a tool that can control Mac applications, remember user preferences, and handle ongoing automated tasks.

The update, announced via LinkedIn, brings several new features to the desktop version of ChatGPT's Codex interface. Users can now direct Codex to interact with applications running on their Mac, connect to a broader range of third-party tools, and generate images directly within the workflow.

The most significant addition is Codex's ability to learn from previous actions and remember how individual users prefer to work. This creates a more personalized experience over time, with the AI adapting its responses and suggestions based on past interactions. The system can also take on repeatable tasks, suggesting OpenAI is positioning Codex as a tool for workflow automation rather than just one-off queries.

Desktop-First Strategy

The features are launching exclusively on the Codex desktop app rather than the web interface, following ChatGPT's recent push into workspace automation. This desktop-first approach mirrors strategies from competitors like Cursor and Windsurf, which have gained traction by offering native desktop experiences with deep system integration.

Mac app control puts ChatGPT in direct competition with automation tools like Zapier and Make, though with a conversational interface rather than visual workflow builders. Users can potentially ask Codex to perform actions across multiple applications without manually creating integration recipes.

The image creation capability suggests integration with DALL-E, OpenAI's image generation model. This would place visual content creation directly within the Codex workflow, eliminating the need to switch between ChatGPT and dedicated image tools like Midjourney.

Memory and Context

Codex's memory feature represents a step toward persistent AI assistants that maintain context across sessions. Rather than starting each conversation from scratch, the system can reference previous work, recall user preferences, and apply learned patterns to new tasks.

This memory capability could prove particularly valuable for repetitive business processes. A user who regularly formats data in a specific way or follows consistent workflows could train Codex through examples, then invoke those patterns with simple commands.

The announcement lacks technical details about how memory functions, what data is stored locally versus in the cloud, or how users can manage or delete stored information. OpenAI has not specified which subscription tiers will have access to these features or whether they will eventually expand beyond Mac to Windows and Linux.

The rollout is described as starting today, suggesting a gradual deployment rather than immediate availability for all users. OpenAI did not provide a timeline for when the features will reach the full user base or announce plans for the web version of ChatGPT.

Sources

1 checked

How we cover tool news: Create With's tool desk drafts these reports with AI from the sources listed above and checks them against those sources before publishing.

Worth passing on?

ShareLinkedIn

Go deeper on ChatGPT

Related reading, watching and going.

Everything on ChatGPT →

The briefing

21 Oct 2026

Stop Asking AI — Build Reusable Scripts to Save Tokens

Discover why asking AI to write small reusable scripts — not one-off replies — saves tokens, improves accuracy and privacy, and makes tools easy to reuse.

Read →

The briefing

13 Oct 2026

Jev - The System One AI That Never Writes a Word, Explained

Deep dive into Jev - TypeSafe's 'System One' model that outputs decisions, not text. Read how it works, real demos (driving), what it's good for and its limits.

Read →

The briefing

30 Sept 2026

Solo Consultant's AI Operating System Built on Markdown

London solo consultant uses an AI personal OS of markdown files—saving ~45% delivery time, boosting revenue 2.4×, and powering 33 community events for ~£250/mo.

Read →
36 min

Podcast

55. We're Back! Conference Recap, Running Events with AI & Self-Healing Apps

James and Kieran return to the Create With Podcast after their summer break to recap the 2026 Create With Conference in Brighton. They discuss how 550 attendees experienced workshops, a charity hackathon where functional apps were built in 30 minutes, and an innovative conference operating system built entirely with AI. The episode covers the latest developments in AI development tools including Anthropic's Fable model, scheduled tasks in Codex, and self-healing applications using PostHog.

Watch now ↗
31 min

Conference

How to Get Your Product Recommended by Google and ChatGPT

Tanya Van Gastel presents the APEX framework for making products discoverable in AI-powered search engines like ChatGPT and Google's AI Overviews. She explains how AI is fundamentally changing product discovery and buyer behavior, then breaks down five critical components: Authority (being cited as a solution), Presence (visibility across search platforms), Expertise (creating high-quality money pages), Endorsements (off-site validation), and Extractability (making content easy for AI to process). The talk emphasizes quality over quantity, with data showing 50% of SaaS SEO revenue comes from fewer than 10 high-performing pages.

Watch now ↗
35 min

Conference

The AI-Native Solopreneur: Build Without a Technical Co-Founder

Effie Bersoux presents a paradigm shift in entrepreneurship where AI enables solo founders to build, validate, and scale businesses without technical co-founders or significant capital. Drawing from Growth Girls' experience building 40+ companies in one year and going from a £4m exit (7 years) to a £40m acquisition target (1 year), she demonstrates how AI breaks the traditional capital-talent-traction loop that disadvantages undercapitalized founders. The talk provides concrete frameworks for AI-native operating models, automation decision-making, and rapid idea-to-market validation.

Watch now ↗

Latest tool news

What else changed this week.

All tool news

The Create With Briefing

Don't watch forty changelogs. Read one email.

Every Tuesday: the tool changes worth knowing, real business use cases, and what's on near you. Free, unsubscribe any time.