Inferensys

Glossary

Science-Based Targets (SBTi)

A validation framework ensuring corporate emission reduction goals align with the Paris Agreement's pathway to limit global warming, requiring specific near-term and long-term net-zero commitments.
Stylish WeWork-like workspace with hot desks and document wall, professional searching through enterprise knowledge base on a mounted ultrawide display, warm industrial pendants overhead.
DEFINITION

What is Science-Based Targets (SBTi)?

A validation framework ensuring corporate emission reduction goals align with the Paris Agreement's pathway to limit global warming, requiring specific near-term and long-term net-zero commitments.

The Science Based Targets initiative (SBTi) is a partnership between CDP, the UN Global Compact, WRI, and WWF that defines and promotes best practices in emissions reductions and net-zero targets in line with climate science. It provides companies with a clearly defined pathway to reduce greenhouse gas (GHG) emissions, ensuring their transition plans are independently validated against the goal of limiting global temperature rise to 1.5°C above pre-industrial levels.

For enterprise AI governance, SBTi validation requires organizations to account for Scope 2 emissions from purchased cloud electricity and Scope 3 emissions embedded in the value chain, including hardware manufacturing and model training. The framework mandates near-term targets covering 5-10 years and long-term targets achieving at least 90% emission reductions before neutralizing residual emissions, moving beyond carbon offsetting toward deep decarbonization of compute infrastructure.

SCIENCE-BASED TARGETS

Core Components of the SBTi Framework

The Science Based Targets initiative (SBTi) provides a structured pathway for companies to set greenhouse gas reduction goals grounded in climate science. The framework mandates specific near-term milestones and long-term net-zero commitments to ensure corporate strategy aligns with the Paris Agreement's 1.5°C trajectory.

01

Near-Term Science-Based Targets

Commitments to reduce Scope 1, 2, and 3 emissions within a 5–10 year timeframe. These targets must cover at least 95% of Scope 1 & 2 emissions and, if Scope 3 is material (over 40% of total emissions), 67% of Scope 3 emissions. The reduction pathway must be consistent with limiting warming to 1.5°C, typically requiring a 4.2% linear annual reduction for Scopes 1 and 2.

4.2%
Minimum Annual Reduction (Scopes 1 & 2)
5-10 Years
Target Timeframe
02

Long-Term Net-Zero Targets

A commitment to reduce emissions to a residual level consistent with a 1.5°C scenario by no later than 2050. This requires a total value chain emission reduction of at least 90% before neutralizing the final, unavoidable residual emissions (≤10% of base year) through permanent carbon removal and storage.

90%+
Required Emission Reduction
2050
Deadline at the Latest
03

Scope 3 Engagement & Portfolio Coverage

For financial institutions, the framework requires specific asset class coverage. The Temperature Rating methodology translates portfolio holdings into an implied temperature rise. Companies with significant indirect emissions must set supplier or customer engagement targets, ensuring that a defined percentage of their value chain partners set their own science-based targets within a fixed period.

67%
Minimum Scope 3 Coverage
04

The Validation & Target-Setting Protocol

A five-step process ensuring rigor:

  • Commit: Submit a commitment letter to the SBTi.
  • Develop: Model targets using SBTi-approved tools like the Sectoral Decarbonization Approach (SDA).
  • Submit: Present targets for official validation against SBTi Criteria.
  • Communicate: Publicly announce approved targets.
  • Disclose: Report emissions annually and recalculate targets every 5 years.
05

Sectoral Decarbonization Approach (SDA)

A specific methodology used for homogeneous, energy-intensive sectors (e.g., power generation, cement, aluminum). The SDA allocates the global carbon budget to individual companies based on their physical intensity convergence relative to a sector-specific pathway. It ensures that a company's intensity target converges to the sector average in 2050, preventing free-riding.

06

Beyond Value Chain Mitigation (BVCM)

While not a substitute for direct reduction, SBTi encourages companies to finance external climate action outside their own value chains. This includes investments in permanent carbon dioxide removal (CDR) and avoided emissions projects. The BVCM recommendation is critical for neutralizing the final ≤10% of residual emissions required for the Net-Zero Standard.

SCIENCE-BASED TARGETS

Frequently Asked Questions

Clear, technical answers to the most common questions about the SBTi framework, net-zero commitments, and how corporate emission reduction targets are validated against climate science.

The Science Based Targets initiative (SBTi) is a partnership between CDP, the United Nations Global Compact, World Resources Institute (WRI), and the World Wide Fund for Nature (WWF) that defines and promotes best practice in emissions reductions and net-zero targets in line with climate science. The SBTi functions as the global validation body, providing companies with a clearly defined pathway to reduce greenhouse gas (GHG) emissions in alignment with the goals of the Paris Agreement—limiting global warming to 1.5°C above pre-industrial levels. The process works through a rigorous five-step framework: a company first commits to set a science-based target, develops a target aligned with SBTi criteria, submits the target for official validation, communicates the approved target publicly, and finally discloses progress annually. The SBTi's Target Validation Team performs a detailed technical assessment against quantitative thresholds, ensuring that near-term targets cover a minimum of 5-10 years and long-term targets achieve a residual emission reduction of at least 90% before neutralization. The initiative distinguishes itself by rejecting voluntary carbon offsets as a substitute for direct emission cuts within a company's value chain, mandating that decarbonization efforts focus on absolute emission reductions first.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.