OpenAI launches LifeVal to measure AI on everyday life tasks
On 2 Nov 2026 OpenAI introduced LifeVal, an evaluation of everyday personal-assistant tasks drawn from time-use surveys and a demographically representative survey of about 3,995 people—covering areas such as childcare, government services, and personal finances. OpenAI says it used LifeVal over the past year to track and drive model gains, and shared example scores on sample tasks.
Verified on OpenAI’s post LifeVal (dated November 2, 2026) and the same-day OpenAI account post: OpenAI introduced LifeVal, an evaluation meant to measure how well models handle everyday tasks people want help with, not only academic knowledge checks.
OpenAI says tasks were shaped by the American Time Use Survey and the Multinational Time Use Study, plus a demographically representative survey of about 3,995 respondents on where people most want a personal assistant. Covered areas include caregiving, health, education, household administration, personal finances, shopping, and accessing government services. Examples include choosing the right kindergarten after a move, organizing flood-prep volunteers, finding a suitable home, watching a game remotely with a grandparent, and meal planning after a new diabetes diagnosis.
OpenAI states it has used LifeVal over the past year to track and drive improvements on browser use, computer use, and connectors—moving ChatGPT beyond chat toward a more agentic personal assistant. A table lists example “Best recorded GPT‑6 Astra score” figures on several tasks (many shown as 100%), alongside sample workflow scores. Those figures are OpenAI’s reported results, not independent IdeaDawn retests.
LifeVal is framed as a research and measurement release for everyday assistant progress; the article does not claim a fully public, reproducible external leaderboard or dataset download.