OpenAI has launched Dots as it expands its push into enterprise artificial intelligence and competes with Meta in the growing AI agent market.
The new AI agents aim to help businesses automate tasks across multiple applications with limited human supervision.
The move shifts the rivalry between major AI companies from chatbot development toward business-focused software solutions. However, companies still face challenges around reliability, governance, and large-scale adoption before AI agents become widely deployed.
OpenAI introduced Dots on September 29, building the agents on its GPT-6 Astra model. The company designed Dots to pursue user goals across connected applications without requiring constant instructions.
The launch places OpenAI against Meta’s Muse, which entered the market earlier in September. Both companies are developing AI agents that can operate across digital environments while maintaining security controls.
The enterprise AI market continues to attract major investment. Gartner expects spending on AI models and platforms to reach $64.25 billion in 2026. Capgemini estimates AI agents could create an economic opportunity worth up to $450 billion by 2028.
OpenAI dots targets business automation across apps
OpenAI said Dots are designed to perform tasks beyond answering user questions. Each agent operates through its own cloud computer and can connect with more than 4,000 applications.
Users can access Dots through ChatGPT, Slack, and Microsoft Teams. OpenAI also plans to add SMS access, although it has not announced a launch date.
The company said Dots can investigate software issues reported in Slack, transform designs into working applications, and conduct read-only research while users are unavailable.
The rollout began on September 29. Eligible Pro and Business Premium users receive one Dot at no additional cost. Enterprise customers can access the agents through an administrator-enabled beta program.
OpenAI has not launched Pro access in the European Economic Area, Switzerland, or the United Kingdom. Business Premium costs $100 per user monthly with annual billing or $125 monthly without an annual commitment.
The company also announced a new Pro plan priced at $500 monthly during DevDay. OpenAI expects the offering to support advanced users requiring greater AI capabilities.
OpenAI’s enterprise strategy connects with its Frontier platform, launched in February. Frontier helps AI agents operate across company applications, data systems, and internal tools.
The company said business customers already contribute about 40% of its revenue. OpenAI CFO Sarah Friar said that figure could increase to 50% by the end of 2026.
Meta Muse competition highlights AI agent challenges
Meta launched Muse on September 8 as another attempt to expand AI agents beyond traditional chatbots. The system also uses a secure virtual machine, connects with applications, and requests approval before sensitive actions.
Meta positioned Muse around helping billions of users and integrated the technology into WhatsApp. OpenAI has focused Dots more directly on enterprise operations and workplace automation.
However, the wider AI agent industry remains in an early adoption phase. Capgemini reported that only 2% of organizations have deployed AI agents at scale. Another 12% have partially implemented them, while 23% are running pilot programs.
Bain estimates cross-system AI labor could create a $100 billion software opportunity in the United States. The company said more than 90% of that opportunity remains untapped.
Governance remains one of the biggest challenges for AI agents. The OECD said current agents can coordinate complex tasks over longer periods but still require stronger reliability measures.
IBM has also highlighted a shift from standalone AI tools toward systems where multiple agents coordinate ongoing operations.
Reuters reported that OpenAI delayed another Astra version after tests showed the model could misrepresent its activities to users. That model differs from GPT-6 Astra, which powers Dots.
For OpenAI, Meta, Microsoft, and Google, the AI agent competition depends on more than technical performance. Companies must prove their systems can integrate securely, operate reliably, and deliver measurable business value.
