Meta Platforms Inc.
AI LabLegal name: Meta Platforms Inc. · HQ: US · Public (META) · Founded 2004
Open-source AI lab and hyperscale infrastructure operator. The Llama model family is released openly; training compute is disclosed in research papers but Scope 1/2/3 emissions are not isolated to AI workloads.
Disclosure Quality
Score
38/100
Transparency Grade
T-4
minimal disclosure
Grades run T-1 (most transparent) to T-5 (least transparent).
Red flags identified (3)
- RF-01 · Total energy consumption not disclosed
- RF-03 · Water withdrawal not disclosed
- RF-04 · AI-specific energy use not disclosed
See the disclosure audit below for details.
Fiscal year 2023. Score computed from OpenButterfly T1–T6 disclosure framework audit. Confidence tier: C (Estimated, Documented Method).
Environmental Metrics
Fiscal year 2026. Each metric is shown at its most recent available year; rows from a different year are marked.
Confidence tiers: A measured and verified · B company-reported · C documented estimate · D uncertain estimate · E modeled · ND not disclosed.
Total Energy Consumption
24.3 TWh
≈ the annual electricity use of 2.3 million US homes
method: known_capacity_model
Source: OpenButterfly pipeline
AI-Specific EnergyFY 2023
Not DisclosedNDchecked 2025-05-23AI training and inference energy not separately disclosed.
Scope 1 Emissions
178 kt CO₂e
≈ the annual emissions of 39,000 passenger cars
method: scope1_ratio_model
Source: OpenButterfly pipeline
Scope 2 Emissions (market-based)
446 kt CO₂e
≈ the annual emissions of 97,000 passenger cars
method: grid_carbon_model
Source: OpenButterfly pipeline
Scope 3 Emissions
13.63 Mt CO₂e
≈ the annual emissions of 3.0 million passenger cars
method: scope3_ratio_model
Source: OpenButterfly pipeline
Water Withdrawal
33.5 GL
≈ 13,000 Olympic swimming pools of water
method: wue_model
Source: OpenButterfly pipeline
Property-Line Noise (estimated)
~57.2 dBA at property line
method: cooling_load_noise_model
Source: OpenButterfly pipeline
Change Over Time
Fiscal years with a value on record. The letter under each year is that figure's confidence tier — a change between a disclosed (A–B) and an estimated (C–E) figure is indicative only, not a measured trend. Modeled noise is excluded: its year axis tracks the model, not the facility.
+22% FY 2023 → FY 2026
+264% FY 2024 → FY 2026
+83% FY 2024 → FY 2026
+50% FY 2024 → FY 2026
Disclosure Audit
Show 5 disclosure checks (T1–T6 framework)
Total electricity consumption disclosed
Renewable capacity (MW) disclosed but not total MWh consumed.
RF-01 (Total energy consumption not disclosed)
Scope 2 market-based emissions disclosed
Reported as zero market-based via renewable procurement.
Total water withdrawal disclosed
Water data not disclosed in 2023 report.
RF-03 (Water withdrawal not disclosed)
AI-specific energy consumption disclosed
AI workload energy not disclosed.
RF-04 (AI-specific energy use not disclosed)
GHG accounting methodology documented
GHG Protocol referenced; limited methodology detail.
Sources
- Meta 2023 Sustainability Report· sustainability report · trust: medium · accessed 2025-05-23
Meta discloses renewable capacity but not total energy in MWh.