| AI wins IMO gold medal | Capabilities | XPT | 2022 | Yes | Yes |
| MATH dataset benchmark | Capabilities | XPT | 2022 | Yes | Yes |
| MMLU benchmark | Capabilities | XPT | 2022 | Yes | Yes |
| QuALITY benchmark | Capabilities | XPT | 2022 | Yes | Yes |
| Date of first publicly known advanced AI | Capabilities | XPT | 2022 | No | No |
| Nick Bostrom affirms existence of AGI by 2030 | Capabilities | XPT | 2022 | No | No |
| ImageNet classification training efficiency | Capabilities | XPT | 2022 | No | No |
| GPT revenue stays under $1b | Adoption | XPT | 2022 | Yes | No |
| Cost of compute of largest AI experiment | Adoption | XPT | 2022 | Yes | Yes |
| US computer R&D spending | Adoption | XPT | 2022 | Yes | Yes |
| Maximum compute used in an AI experiment | Adoption | XPT | 2022 | Yes | No |
| Lowest price of 1 GFLOPS | Adoption | XPT | 2022 | Yes | Yes |
| Largest number of parameters in an ML model | Adoption | XPT | 2022 | No | No |
| US GDP from software and information services | Impact | XPT | 2022 | Yes | Yes |
| OECD labor force participation rate | Impact | XPT | 2022 | Yes | Yes |
| Probability of GDP growth over 15% | Impact | XPT | 2022 | No | No |
| Negative public opinion of AI | Impact | XPT | 2022 | Yes | Yes |
| AI catastrophic risk | Impact | XPT | 2022 | No | No |
| AI extinction risk | Impact | XPT | 2022 | No | No |
| NYT bestsellers written by AI | Impact | XPT | 2022 | No | No |
| AI out-forecasts a team of superforecasters | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | Yes | Yes |
| AI solves a famous open math problem | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| AI causes 10,000+ deaths or $10b+ in damage | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Two AI-caused catastrophes of escalating severity | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Global corporate AI investment falls below 25% of its 2021 level | Adoption | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Evidence of AI self-replication and shutdown evasion | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| US real GDP growth averages 4% or more (2023–2030) | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| 100x drop in compute for state-of-the-art AI within one month | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| US implements 4+ of Muehlhauser's 12 AI policy ideas | Governance | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| US requires testing of AI systems before deployment | Governance | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Enforced six-month international pause on AI development | Governance | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| LLM that verifiably never describes violence | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Unprompted AI writes and deploys improved AI software | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Peer-reviewed demonstrations of convergent instrumental subgoals | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| A top-four frontier AI lab is based in a non-democracy | Governance | Roots of Disagreement on AI Risk | Apr–May 2023 | No | Yes |
| AI solely authors top-journal review articles and top-10 apps | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Frontier AI model sustains deception of human experts | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| IT and communications exceed 70% of the S&P 500 | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| AI autonomously influences a democratic election result | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Global real GDP exceeds 115% of any prior year's peak | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Two AI-caused catastrophes of escalating severity (platform version) | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| US implements 5+ of Muehlhauser's 12 AI policy ideas | Governance | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| ARC Evals finds AI able to self-replicate and evade shutdown | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Robot assembles a scale-model car from written instructions | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| AI is the "most important problem" for 15%+ of Americans | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Twofold partisan gap over AI as a top US policy priority | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| US–China intergovernmental AI safety organization | Governance | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| Developers shut down an AI system after power-seeking behavior | Impact | Roots of Disagreement on AI Risk | Apr–May 2023 | No | No |
| AI reliably writes bug-free code of 10,000+ lines | Capabilities | Roots of Disagreement on AI Risk | Apr–May 2023 | Yes | Yes |
| VCT: probability that AI matches the median virologist | Capabilities | LLM Biorisk | Late 2024 / early 2025 | Yes | Yes |
| VCT: AI matches or outperforms experts | Capabilities | LLM Biorisk | Late 2024 / early 2025 | Yes | Yes |
| AI outperforms experts on long-form biothreat questions | Capabilities | LLM Biorisk | Late 2024 / early 2025 | Yes | Yes |
| Probability of large-scale human-caused epidemic conditional on AI capabilities | Impact | LLM Biorisk | Late 2024 / early 2025 | No | No |
| Influenza synthesis: AI enables 10% of non-experts to synthesize influenza virus | Capabilities | LLM Biorisk | Late 2024 / early 2025 | Yes | Yes |
| Influenza synthesis: AI enables 50% of non-experts to synthesize influenza virus | Capabilities | LLM Biorisk | Late 2024 / early 2025 | No | No |
| Bioweapon attack planning: AI enables plausible plan | Capabilities | LLM Biorisk | Late 2024 / early 2025 | Yes | Yes |
| Dual-use DNA: AI enables 90% success rate | Capabilities | LLM Biorisk | Late 2024 / early 2025 | Yes | Yes |
| AI Cyber: year AI solves >90% of tasks on Cybench | Capabilities | AI Cyber | 2025 | Yes | Yes |
| AI Cyber: year AI enables 25% of moderately skilled individual hackers to find vulnerabilities and write elite exploits | Capabilities | AI Cyber | 2025 | Yes | No |
| AI Cyber: year AI enables novice hackers to match an elite team in an operational technology capture-the-flag-style competition | Capabilities | AI Cyber | 2025 | Yes | Yes |
| AI Cyber: AI-enabled actor first causes a $100m+ grid blackout | Impact | AI Cyber | 2025 | Yes | Yes |
| Active Site RCT: forecasting internet-only arm | Capabilities | Active Site RCT | 2025 | No | No |
| Active Site RCT: forecasting internet-plus-LLM arm | Capabilities | Active Site RCT | 2025 | Yes | Yes |
| State-of-the-art AI accuracy on FrontierMath benchmark | Capabilities | LEAP W1 | Jun–Aug 2025 | Yes | Yes |
| Autonomous vehicles: share of ride-hailing trips | Adoption | LEAP W1 | Jun–Aug 2025 | Yes | Yes |
| Change in US jobs by occupational class | Capabilities | LEAP W1 | Jun–Aug 2025 | Yes | No |
| General AI progress | Impact | LEAP W1 | Jun–Aug 2025 | No | No |
| AI's impact on the Technological Richter Scale | Impact | LEAP W1 | Jun–Aug 2025 | No | Yes |
| AI solves or substantially assists a Millennium Prize Problem | Capabilities | LEAP W2 | Aug–Sep 2025 | Yes | Yes |
| Share of science publications that are AI-engaged | Adoption | LEAP W2 | Aug–Sep 2025 | No | Yes |
| Share of new drug sales from AI-discovered drugs | Adoption | LEAP W2 | Aug–Sep 2025 | Yes | Yes |
| Share of US electricity used to train and run AI | Adoption | LEAP W2 | Aug–Sep 2025 | Yes | Yes |
| Annual global private AI investment | Adoption | LEAP W3 | Sep–Oct 2025 | No | No |
| Share of US work hours assisted by generative AI | Adoption | LEAP W3 | Sep–Oct 2025 | Yes | Yes |
| Benchmark gap: open- vs. closed-weight frontier models | Capabilities | LEAP W3 | Sep–Oct 2025 | Yes | Yes |
| Share of US adults using AI for companionship daily | Adoption | LEAP W3 | Sep–Oct 2025 | Yes | Yes |
| Annual recurring revenue of the top AI company | Adoption | EEAI | Oct 2025–Feb 2026 | Yes | Yes |
| Change in GDP | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Change in labor productivity | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Change in total factor productivity | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Labor force participation rate | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Unemployment rate | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Youth unemployment rate | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Share of job categories | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Change in employment by occupation | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Wealth inequality | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Labor share | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Median household income | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Life satisfaction | Impact | EEAI | Oct 2025–Feb 2026 | No | No |
| Work hours assisted by Gen AI | Adoption | EEAI | Oct 2025–Feb 2026 | No | No |
| AI electricity consumption | Adoption | EEAI | Oct 2025–Feb 2026 | No | No |
| State-of-the-art accuracy on LiveCodeBench Pro Hard | Capabilities | LEAP W4 | Nov–Dec 2025 | Yes | Yes |
| Valuation per employee at the top five AI companies | Adoption | LEAP W4 | Nov–Dec 2025 | Yes | No |
| Entry-level share of new hires at top 15 tech companies | Impact | LEAP W4 | Nov–Dec 2025 | No | Yes |
| Year three $10b companies have five or fewer employees | Impact | LEAP W4 | Nov–Dec 2025 | Yes | No |
| Global hyperscale data center capacity | Adoption | LEAP W4 | Nov–Dec 2025 | Yes | Yes |
| AI-coauthored AI paper judged worthy of a Test-of-Time award | Impact | LEAP W4 | Nov–Dec 2025 | No | No |
| US–China frontier-model gap on the Epoch Capabilities Index | Capabilities | LEAP W5 | Jan–Feb 2026 | Yes | Yes |
| Manufacturing origin of frontier AI training chips | Governance | LEAP W5 | Jan–Feb 2026 | No | Yes |
| NATO authorization of fully autonomous cyber weapons | Governance | LEAP W5 | Jan–Feb 2026 | No | Yes |
| Regulatory fines on AI-generated content | Impact | LEAP W5 | Jan–Feb 2026 | No | No |
| ForecastBench: year AI beats superforecasters | Capabilities | LEAP W5 | Jan–Feb 2026 | Yes | No |
| US real GDP growth (annualized) | Impact | LEAP W6 | Feb–Mar 2026 | Yes | Yes |
| US labor force participation rate | Impact | LEAP W6 | Feb–Mar 2026 | Yes | Yes |
| Wealth inequality: top 10% share of US wealth | Impact | LEAP W6 | Feb–Mar 2026 | Yes | Yes |
| Industrial robots in use globally | Adoption | LEAP W7 | Mar 2026 | No | Yes |
| US labor share | Impact | LEAP W7 | Mar 2026 | No | No |
| Year a robot passes the Coffee Test in an unfamiliar home | Capabilities | LEAP W7 | Mar 2026 | No | Yes |
| Year a robot performs an appendectomy without physical human help | Capabilities | LEAP W7 | Mar 2026 | No | No |
| Amazon e-commerce sales per distribution employee | Adoption | LEAP W7 | Mar 2026 | No | No |
| AI's impact on the Technological Richter Scale (second asking) | Impact | LEAP W8 | Apr–May 2026 | No | Yes |
| Year a majority of LEAP panelists agree AGI exists | Impact | LEAP W8 | Apr–May 2026 | No | No |
| State-of-the-art METR 80% time horizon | Capabilities | LEAP W8 | Apr–May 2026 | Yes | Yes |
| METR 80% time horizon: AI completes 8-hour tasks | Capabilities | LEAP W8 | Apr–May 2026 | Yes | No |
| Rapid progress scenario timelines | Impact | LEAP W8 | Apr–May 2026 | No | No |
| AI's overall impact on the US over 20 years | Impact | LEAP W8 | Apr–May 2026 | No | No |
| Year the US, UK, or EU requires pre-approval of frontier AI models | Governance | LEAP W9 | May–Jun 2026 | Yes | Yes |
| Cybercrime losses reported to the FBI | Impact | LEAP W9 | May–Jun 2026 | No | Yes |
| AI catastrophic risk | Impact | LEAP W9 | May–Jun 2026 | No | No |
| First major AI global harm event | Impact | LEAP W9 | May–Jun 2026 | No | No |
| Top-10 economies rated "authoritarian" on the EIU Democracy Index | Impact | LEAP W9 | May–Jun 2026 | No | No |
| Deaths from non-eradicable diseases | Impact | LEAP W10 | Jun–Jul 2026 | No | No |
| US life expectancy | Impact | LEAP W10 | Jun–Jul 2026 | No | Yes |
| Global real GDP growth | Impact | LEAP W10 | Jun–Jul 2026 | No | No |
| Consumer value of generative AI | Adoption | LEAP W10 | Jun–Jul 2026 | No | No |
| Global average happiness | Impact | LEAP W10 | Jun–Jul 2026 | No | No |
| Frontier AI company independence | Impact | LEAP W11 | Jul–Aug 2026 | No | No |
| Frontier AI company revenue | Adoption | LEAP W11 | Jul–Aug 2026 | Yes | Yes |
| AI-exposed equity performance: software sector | Adoption | LEAP W11 | Jul–Aug 2026 | No | No |
| AI-exposed equity performance: semiconductor sector | Adoption | LEAP W11 | Jul–Aug 2026 | No | No |
| Data center investment | Adoption | LEAP W11 | Jul–Aug 2026 | No | No |