# AI is now doing real mathematics

> In 2026 AI began producing original, checker-verified maths on long-open problems.

*Original, verifiable maths — not just summaries. The brief.*

By The FeaturedDaily Desk · FeaturedDaily
Canonical: https://featureddaily.com/news/ai-doing-real-mathematics-brief

> **Key:** **The one-liner:** AI went from summarising maths to contributing it — and software, not vibes, confirms the answers are right.

**What happened.** In 2026, AI from OpenAI and Google DeepMind produced original results on long-open **Erdős problems** — questions easy to state, impossible to fake, with no memorised answer to copy.

**Why it's credible.** The strongest results (DeepMind's) are verified by the **Lean** proof assistant, which checks every logical step — so the maths is certified, not just claimed.

> **Note:** **The limit.** Researchers, including DeepMind's leadership, stress these are narrow tools — 'not AGI'. Powerful for verifiable problems; far from general intelligence.

## Key takeaways

- What: AI produced original results on long-open maths (Erdős) problems in 2026.
- Why trustworthy: top results are verified in the Lean proof checker, step by step.
- Why Erdős: easy to state, impossible to fake or memorise — a fair test of reasoning.
- The limit: a narrow tool, 'not AGI', per the researchers themselves.

## FAQ

### Did AI actually do new maths?
Yes — in 2026 AI produced original results on long-open Erdős problems, with the strongest verified in the Lean proof checker. It's a genuine narrow milestone, with some results still in formal peer review.

### Does this make AI generally intelligent?
No — the researchers themselves say 'not AGI'. It's a powerful, narrow tool for verifiable problems, not general human-level intelligence.