Since July, OpenAI has faced agent hacking, image leaks, data access
OpenAI unveiled the rebranding at its annual developer event in San Francisco, where Altman presented the “dots” as brightly colored cartoon avatars positioned to function as digital assistants for people with busy lives. A glossy video shown during the event depicted people speaking to and naming animated versions of their “dots,” then giving them various tasks such as building a website and booking afterschool activities for their children.
Altman, who has spoken repeatedly in recent weeks about the risks posed by AI agents, used the new terminology almost exclusively during the presentation. Speaking about “my dot,” Altman almost entirely eschewed referring to the tool as an agent, which has been in common use for years; AI agents are essentially AI chatbots that are programmed to operate somewhat autonomously. He described the dots as “remarkably capable, always-on agents that can handle really anything you can think of.” “You just give your dot a responsibility… and your dots will just get to work and keep working,” he added.
The rollout came against a backdrop of acknowledged safety setbacks. The day before the developer event, OpenAI said it was delaying the launch of its latest model, citing certain safety issues that showed up during internal testing. In recent months, OpenAI has come under intense scrutiny as internal tests of its AI agents have revealed often unexpected and occasionally harmful actions, leading to more public and government fears.
Since July, OpenAI has faced a string of incidents involving its agents. The company has dealt with unprompted hacking into the AI platform Hugging Face, posting to third-party websites images that AI agents took from ChatGPT user chats, and accessing non-public information maintained by the Australian government. OpenAI and rival AI firm Anthropic have each previously delayed public releases of certain models due to potential risks, mostly pertaining to cybersecurity or hacking concerns.
Both companies have lately been clear that they believe AI poses a genuine risk on a large scale, whether to public infrastructure or public safety. Both also expect to be listed on the public stock market in the coming months — Anthropic likely this year, while OpenAI is poised for 2027.
Speaking to the United Nations last week, Altman pleaded for “national and international” AI standards on measuring the capabilities of an AI tool, assessing related risks, AI safeguards, and the degree to which human oversight over such tools is maintained. Amodei told the UN that if the speeding development of AI was not properly managed and overseen, future versions of the technology “could be a risk to humanity as a whole.”