# AI Reconstructs Software From Behavior Alone

AI agents increasingly reconstruct working software by observing its behavior without source access, and new benchmarks track this fast-rising capability — eroding source-code secrecy as a moat and raising IP and security implications as more such results accrue.

- Conviction: 29 / 100 (weakening)
- 7-day move: -7
- Horizon: Emerging (watchlist)
- Tracking since: 2026-06-28T00:00:00.000Z
- Last updated: 2026-07-07T14:00:02.329Z
- Canonical: https://polylog.news/ai/trends/ai-software-reconstruction-from-behavior
- Publisher: Polylog
- Affected regions: Global

## Recent score history

- 2026-07-06: 30
- 2026-07-07: 29

## Recent evidence

- [confirms] In a New Benchmark, an AI Agent Rebuilds a 60,000-Line Program From Its Behavior Alone (2026-06-28): Epoch AI and METR released MirrorCode, a benchmark measuring whether models can rebuild a 60,000-line program from its behavior alone; early scores ran higher than program-synthesis researchers expected.
- [confirms] MirrorCode Benchmark Finds Frontier Models Can Rebuild Real Software From Scratch, Up to a Point (2026-06-28): Epoch AI and METR's MirrorCode benchmark tasks models with rebuilding working programs without source code, finding a best 56 percent solve rate on tasks estimated to take human engineers weeks.