Gap between top US labs narrows to 0-2 months

What AI 2027 Predicted

The AI 2027 scenario depicted a single dominant lab (“OpenBrain”) maintaining a 3–9 month lead over competitors. In their own February 2026 self-grading, the AI Futures Project authors acknowledged this prediction was wrong: “The race appears to be closer than we predicted, more like a 0-2 month lead.” The frontier is more competitive than the scenario anticipated.

How We Track This

We monitor:

• Release dates of frontier models from OpenAI, Anthropic, and Google DeepMind
• Benchmark performance at time of release (Chatbot Arena, LMSYS, coding benchmarks)
• Leapfrogging patterns — how quickly one lab’s lead is matched
• xAI (Grok) as a potential fourth frontier competitor
• Qualitative assessments from researchers and industry observers

Current Evidence

AI Futures Self-Grading (Feb 2026): The scenario’s own authors explicitly graded this prediction as incorrect. They found the race “closer than predicted, more like a 0-2 month lead” rather than the 3-9 months depicted.

• Grading AI 2027’s 2025 Predictions — AI Futures Project

November–December 2025 Release Cluster: An unprecedented concentration of frontier model releases occurred within a ~3 week window: GPT-5.2, Claude 4.5, Gemini 3, and Grok 4.1 all launched between November 17 and December 11, 2025. No single lab maintained a lead for more than days before being matched.

• AI Model Releases Nov/Dec 2025 — VERTU

Four-Way Competition: Unlike AI 2027’s depiction of a single dominant lab, the frontier is contested by at least four major players (OpenAI, Anthropic, Google DeepMind, xAI), with each regularly leapfrogging on different benchmarks. This multi-polar competition dynamic was not anticipated in the scenario.

Benchmark Parity: On major benchmarks (Chatbot Arena, SWE-bench, math competitions), the top models from different labs typically score within a few percentage points of each other. Leadership on any given benchmark frequently changes with each new release.

Counterevidence & Limitations

• Benchmark parity does not necessarily mean capability parity. Labs may have different strengths (Anthropic in safety/coding, Google in multimodal, OpenAI in consumer products).
• Internal capabilities at labs may still have significant gaps even if public releases are close. AI 2027’s prediction was partially about internal capabilities, not just public releases.
• The tight public competition could be partly strategic — labs may be holding back releases to time them against competitors rather than reflecting actual development pace.
• The 0-2 month assessment comes from the AI Futures authors themselves, who may be calibrating against their own scenario generously.
• xAI’s emergence as a frontier competitor was not anticipated in the scenario, adding a dimension of competition the authors didn’t model.

What Would Change Our Assessment

• Maintain “confirmed”: Continued pattern of rapid leapfrogging, no lab sustaining a lead beyond 1-2 months
• Potential revision: If one lab (especially OpenAI) pulls clearly ahead with a next-generation model that takes 6+ months to match, the original 3-9 month gap prediction would be partially vindicated
• Watch for: Whether the gap opens up as compute scales — the lab with the biggest training cluster may eventually pull ahead

Update History

Date	Update
2025-04	Meta Llama 4 (April 5, Apache 2.0) and Alibaba Qwen3 (April 28) both produce open-weight models competitive with some closed frontier models. The competitive surface for frontier capabilities is widening beyond the top three US labs.
2025-05	Claude Opus 4 (May 22), Google Gemini 2.5 Flash GA (May 20), and OpenAI’s continued iteration: three major labs releasing competitive models within days of each other. The gap between top US labs is compressing.
2025-08	GPT-5 (August 7), Gemini 2.5 Deep Think availability expansion (August 1), and Claude Opus 4.1 (August 5) all release within days of each other. Multiple frontier-class releases in the same week from competing labs.
2025-11	Gemini 3 and GPT-5.1-Codex-Max release on November 18; Claude Opus 4.5 releases on November 24 — six days later. Labs are trading benchmark leadership on a weekly basis. The AI 2027 prediction of 0-2 month gap between top labs appears confirmed.
2025-12	GPT-5.2 released three weeks after Gemini 3 triggered an internal “Code Red” at OpenAI. The reactive cadence — Gemini 3 (Nov 18) → GPT-5.2 (Dec 11) — illustrates a gap of weeks, not months, between top US labs.
2026-01	AI Futures Project authors grade the inter-lab gap as 0-2 months, tighter than the 3-9 months predicted in AI 2027. AI Futures Project’s clarification post confirmed: “The race appears to be closer than we predicted, more like a 0-2 month lead between the top US AGI companies.”
2026-02	AI Futures grading explicitly conceded: “The race appears to be closer than we predicted, more like a 0–2 month lead between the top US AGI companies.” The scenario’s 3–9 month OpenBrain lead did not materialize.
2026-03	Frontier race between OpenAI, Anthropic, and Google is extremely tight. The corrected 0-2 month gap is clearly borne out by alternating benchmark leads.