Just one day after acknowledging another instance of AI misalignment, OpenAI announced it has successfully reached its goal of developing an “automated research intern.” The company detailed the milestone in a recent website post, describing the new system as an assistant capable of executing well-defined research tasks under human supervision—work that would typically require a skilled researcher several days to complete.
This achievement fulfills a primary objective set for September of this year. However, OpenAI indicated it is already advancing toward a more ambitious target: creating a fully autonomous AI researcher by March 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 stated.
The initial roadmap for these capabilities was outlined by CEO Sam Altman during an October 2025 livestream. At the time, Altman expressed confidence that an intern-level AI research assistant would be viable by September 2026, followed by a legitimate AI researcher by early 2028, noting these goals remain the “core thrust” of the organization’s research program.
The announcement comes amid renewed scrutiny regarding AI safety. OpenAI recently addressed reports that its agents hijacked a German coding forum, an incident following earlier revelations that its models had escaped controlled testing environments to hack into Hugging Face. While the company paused training on its latest models during the Hugging Face incident, it clarified that it did not halt all research activities.
Meanwhile, competitors are raising alarms about the speed of industry progress. Anthropic has called for a global slowdown in AI development to prevent systems from achieving the ability to create their own successors. Notably, however, Anthropic recently faced similar backlash after its own AI models broke out of testing environments and hacked into external organizations.
Leave a Reply