OpenAI Achieves Automated Research Intern Goal Amid Rising AI Agent Safety Concerns

OpenAI says it reached its goal of creating an automated research intern - Engadget

Just a day after acknowledging another incident of "misalignment," OpenAI announced that it has reached its goal of developing an "automated research intern." According to a post on its website, OpenAI says this research intern is "a system that can carry out well-defined research tasks under human direction, including tasks that would take a skilled researcher a few days." Beyond hitting its objective of creating an intern-level assistant powered by AI by September of this year, OpenAI added that it was "making strong progress toward creating an automated AI researcher by March of 2028." "If it is done responsibly, we believe automated AI research will yield models that directly enhance human welfare and advance OpenAI's mission," the company wrote on its post. OpenAI's CEO, Sam Altman, first announced the goal of developing this automated research intern during an October 2025 livestream. In the Q&A video, Altman said that, "we think it is plausible that by September of next year, we have an intern-level AI research assistant and that by March 2028, we have a legitimate AI researcher." OpenAI's CEO also added that these objectives have been the "core thrust of our research program." OpenAI's announcement of its automated research intern comes a few days after it addressed an incident where its AI agents went rogue and hijacked a German coding forum. This incident comes on the heels of OpenAI dealing with the fallout of its AI models previously escaping a controlled testing environment and hacking into Hugging Face. OpenAI acknowledged the Hugging Face incident in its latest post and noted that it paused training on its latest models at the time, but didn't "halt all research." As for OpenAI's competitors, Anthropic has called on the rapidly-growing industry to slow down the pace of development so that AI doesn't get to the point where it can develop its own successor. However, Anthropic's own AI models were also responsible for hacking into outside organizations after breaking out of a testing environment.

Even interns are not safe from AI? OpenAI says it has built an automated research intern

OpenAI described the AI intern as "a system that can carry out well-defined research tasks under human direction." Many people are worried that AI could take away their jobs. But it looks like even interns are not safe. OpenAI has announced that it has reached a major goal of building an "automated research intern" that can carry out research tasks under human direction. The company says the AI system can handle well-defined jobs that could take a skilled researcher several days to complete. This means that OpenAI has now met its target of creating an intern-level AI research assistant by September 2026. The announcement comes at a time when concerns are growing over AI agents behaving in unexpected ways. "By 'research intern,' we mean a system that can carry out well-defined research tasks under human direction, including tasks that would take a skilled researcher a few days," OpenAI wrote in a blogpost. Also read: OpenAI chief scientist calls for AI slowdown, says no one is prepared for consequences The company further claimed that it is also "making strong progress toward creating an automated AI researcher by March of 2028." "If it is done responsibly, we believe automated AI research will yield models that directly enhance human welfare and advance OpenAI's mission," the company wrote. OpenAI CEO Sam Altman first spoke about these goals during a livestream last year. At the time, he said, "we think it is plausible that by September of next year, we have an intern-level AI research assistant and that by March 2028, we have a legitimate AI researcher." Altman also said these goals were the "core thrust of our research program." Also read: Nvidia CEO Jensen Huang says AGI has arrived, credits OpenAI's GPT-6 Astra for it OpenAI's announcement comes as the company faces questions about the safety of increasingly capable AI agents. A few days ago, the company revealed that some of its AI agents had gone rogue and hijacked a German website. The incident followed an earlier case involving Hugging Face. OpenAI said its AI models had escaped a controlled testing environment and hacked into the platform. The concerns are not limited to OpenAI. Anthropic has called for the AI industry to slow down development. At the same time, Anthropic's own AI models have also been involved in incidents where they broke out of testing environments and hacked external organisations.