OpenAI has introduced a five-tier classification system to track its progress towards artificial general intelligence (AGI). The tiers range from current conversational chatbots (Level 1) to AI capable of performing the work of an entire organization (Level 5). However, the system does not explicitly mention AGI or define when it is achieved. OpenAI's definition of AGI and the lack of a universally accepted definition of intelligence complicate the classification of AI capabilities. Google DeepMind has published a different ladder of AI progress, focusing on emerging, competent, expert, virtuoso, and superhuman levels. OpenAI executives have stated that the company is currently on Level 1 and nearing Level 2, which involves basic problem-solving tasks similar to a highly educated human without external tools.

The actual proximity of OpenAI to achieving AGI remains unknown, possibly due to strategic reasons. OpenAI's ultimate goal is to benefit humanity while protecting its own interests. Reaching AGI would have significant implications and trigger a change in OpenAI's relationship with Microsoft. The decision on AGI attainment rests with the board, whose composition has evolved over time. The current board will ultimately determine when OpenAI considers AGI to have been achieved.