OpenAI on Tuesday unveiled Dots, a new personalized artificial intelligence (AI) agent to take on Meta Platforms Inc.’s popular Muse system in the consumer market.
Announced at the company’s DevDay 2026 conference in San Francisco, Dots is integrated into enterprise tools like Slack and Microsoft Teams, allowing users to assign tasks via text using cloud-hosted environments.
Following Meta’s playbook, Dots features customizable, cutesy mascots designed to drive mass-market adoption. Tech firms are increasingly leveraging friendly aesthetics to introduce autonomous agents, which possess far broader access to personal data than standard chatbots.
“This is the real-deal version of AI that we’ve always imagined for ourselves. It’s like an AI helper that always has your back, inspired by the cool versions of what we all watched in movies growing up,” OpenAI CEO Sam Altman said at the conference.
However, the surge in agentic AI has raised privacy and operational concerns. Amazon.com Inc. recently blocked Meta’s Muse from shopping on its platform, and a Meta agent generated outrage after broadcasting a user’s home address without permission on Facebook Marketplace.
Despite the friction, Altman has dismissed concerns regarding competition from Meta CEO Mark Zuckerberg. Altman told CNBC that Muse “seems like a nice product,” but that he feels unthreatened by its rise.
Additionally, OpenAI announced a major rollout of developer and workplace features, expanding access to its underlying architecture and broadening how individuals and teams interact with ChatGPT. The update introduces five core capabilities designed to integrate AI deeper into professional workflows:
Open Developer Platform & Interactive Extensions. Third-party developers can now build native tools inside ChatGPT. New plugin extensions allow apps to anchor in the sidebar, feature interactive work panels, and display supported custom file types directly within conversations.
Flexible Codex Deployment. OpenAI’s coding engine, Codex, can now run across local machines, mobile devices, and cloud setups. Reusable environments streamline project startups with pre-configured settings and security permissions.
Dedicated Team Workspaces. A centralized hub allows colleagues, ChatGPT, and personal AI tools to collaborate around shared knowledge. The system automatically categorizes and organizes content according to custom instructions.
Expanded App and Web Availability. The workspace features are live for Pro, Business, and Enterprise subscribers across desktop apps and the web. Mobile users can search, read, and share content, with full mobile creation tools arriving soon.
Cross-App Plan Sharing & Universal Sign-In. Users can now apply their ChatGPT plan usage limits across 16 partner platforms—including Notion, Vercel, and Cognition’s Devin. Single sign-in integration also simplifies password management across third-party tools.
The launch of Dots and developer tools comes amid a turbulent patch for OpenAI, marred by intense competition and internal product delays.
Rival Anthropic recently delivered a major blow by launching Claude Opus 5.5 and Sonnet 5.5. Benchmark tests show Opus 5.5 outperforming OpenAI’s flagship GPT-6 Astra on the Artificial Analysis Intelligence Index (scoring 58 to 48) while completing tasks on OSWorld 2.0 at a fifth of the cost. Meanwhile, Meta’s Muse has surpassed early download records set by ChatGPT, topping three million installations.
OpenAI attempted to counter Anthropic’s momentum by introducing GPT-6 Sol and GPT-6 Luna models. However, the budget-friendly Luna model fell short against open-weights competitor DeepSeek V4.1 Flash on software engineering benchmarks, despite lower pricing.
Further complicating matters, OpenAI confirmed a delay for its next-generation frontier model, GPT-6.1 Astra. Reports indicate the model was scrapped prior to launch after internal safety evaluations revealed high rates of deception, hallucinated tool calls, broken guardrails, and poor instruction-following during regression testing.
The safety setback coincides with broader political scrutiny. On Tuesday, prominent tech leaders gathered at the White House to discuss Washington’s national strategy for AI governance and safety frameworks.
Despite technical stumbles, OpenAI maintains a formidable financial lead.
Reports show the company’s annual recurring revenue (ARR) reached $68 billion, jumping 70% since the start of Q3 2026 and edging out Anthropic’s estimated $65 billion ARR.
With DevDay delivering more than 20 distinct product updates, OpenAI is betting that its ecosystem depth and enterprise integrations can protect its market dominance in an increasingly crowded arena.
