Artificial general intelligence will not be achieved in 2025.

Judge at the end of 2025. TRUE: no AI system released or demonstrated in 2025 is regarded as artificial general intelligence by broad expert consensus. Claims by a single company or executive, or marketing language, do not count as achievement. FALSE: by the end of 2025 there is broad expert and community consensus that some system achieved AGI during 2025. MOSTLY FALSE: a 2025 system is widely, though not by consensus, accepted as AGI, with substantial expert support and independent evaluations backing the claim. This is a single-outcome prediction, so MIXED does not apply unless the expert community is genuinely evenly divided.

Set before the research began.

1. E1 ARC Prize 2025 Results and Analysis | ARC Prize ARC Prize · 5 December 2025

   “So, do we have AGI? Not yet.”

2. E2 No AGI by 2025 — Gary Marcus predictionscorecard.com

   “No system recognized as AGI appeared in 2025; by late 2025 even leading figures such as Ilya Sutskever said scaling had stalled and new research was needed, putting human-level learning years away.”

3. E3 2025: The Year in Review gonzoml (Substack) · 6 January 2026

   “The AGI/ASI hype seems to have deflated somewhat.”

4. E4 Status of Artificial General Intelligence (AGI): October 2025 ETC Journal · 17 October 2025

   “As of October 17, 2025, artificial general intelligence (AGI) remains a rapidly evolving but still unachieved goal.”

ARC Prize's 2025 results analysis, published on 2025-12-05, asks whether AGI has arrived and answers "Not yet" [E1]. An independent prediction tracker reports that no system recognized as AGI appeared in 2025 [E2]. A year-in-review published in January 2026 says the AGI hype had deflated somewhat [E3]. None of the evidence shows broad expert agreement that any system reached AGI during 2025, so the prediction resolves true.