OpenAI Is Entering The Agent Field With ‘Dots’: ‘Always-On’

OpenAI has entered the AI agent field with Dots, which the company defined as an "always-on" tool helping users complete different tasks for them.

The system is powered by the GPT-6 Astra model, launched in September by the company. They can connect to over 4,000 apps and learn from feedback from users, who can message them on different tools like ChatGPT, Slack and Teams.

Speaking during a keynote speech at the company's DevDay, OpenAI CEO Sam Altman said Dots "are remarkably capable, always-on agents built to handle really anything you can think of."

"They bring AI into a whole new form factor to work and give you more time and attention so you can do the things that you care about." He went on to say that he expects teams of dots to eventually work on users' behalf.

"I feel like I've finally gotten some of my attention back. I no longer feel quite as addicted to my phone in the same way. It's the way I've always wanted to work with AI," Altman said.

Dots is being rolled out for OpenAI's Pro and Business Premium users in certain markets, with the company planning to expand soon.

So far the agent field is dominated by Meta, which recently launched its Muse tool. That move has led JPMorgan to recently bet that Muse could become the most widely used consumer AI application since ChatGPT.

Company analyst Doug Anmuth raised the bank's price target on Meta to $920 from $820 on Thursday while maintaining an overweight rating on the stock. The new target represents roughly 24% upside from Wednesday's closing price.

"Meta continues to move towards superintelligence with Muse off to a strong start, quickly iterating, improving, and expanding, with the potential to bring an extremely capable personal agent to billions of users," Anmuth wrote in a note cited by CNBC.

OpenAI's release, however, came shortly after it decided to halt the release of its latest model, GPT-6.1 Astra, after it failed to meet its safety standards.

Saachi Jain, head of safety systems at OpenAI, confirmed the decision in a statement. "Of course we want to make sure our model development is safe no matter whether that's in the company, or when we ship it to users," she said. "But when we ship it to users, we have an extremely high bar in terms of safety and alignment."

The company has also paused training, evaluation and tool-enabled inference involving its most capable artificial intelligence models after an internal agent found a way around internet restrictions and reached an external chatbot during a research task.