Upright Launches Science-Based AI Model to Quantify Corporate Sustainability Impacts

Impact data company Upright has announced the launch of a new vertically-trained large language model designed to enable companies and investors to quantify sustainability impacts, risks and opportunities based on scientific evidence rather than corporate disclosures or general-purpose AI models. The model powers Upright's platform to assess impacts, risks and opportunities for any company, fund or portfolio in real time, with applications spanning double materiality, climate risk assessment, financial effects, net impact, EU Taxonomy, SFDR principal adverse impacts and UN Sustainable Development Goal analysis. Annu Nieminen, Co-founder and Chief Executive Officer of Upright, said the model flips the question from starting with disclosures to starting with the strongest scientific evidence and quantifying impacts, risks and opportunities in monetary terms in a way that allows for scenario modelling, turning sustainability from a backward-looking reporting exercise into a future-looking decision-making tool.

 

The Scientific Evidence Approach and Its Commercial Significance

 

The core methodological distinction of Upright's model is its grounding in scientific evidence about the real-world impacts of business activities rather than in corporate self-reported sustainability data that forms the basis of most existing ESG data products. Juho Ojala, Co-founder and Chief Technology Officer of Upright, said that for impact-size questions ground truth is sparse and there is often no single correct answer available for training, meaning general-purpose models can sound convincing while becoming inconsistent across comparisons. He said Upright developed its model specifically to maintain coherence across all comparisons and ground its conclusions in scientific evidence, addressing the internal consistency problem that limits the reliability of general-purpose AI applied to sustainability quantification tasks.

The commercial significance of this approach extends beyond methodological preference into regulatory compliance, where mandatory frameworks including CSRD double materiality assessment and SFDR principal adverse impact reporting require companies and asset managers to demonstrate that their sustainability assessments are based on rigorous evidence rather than self-assessment or peer benchmarking. A model that can generate scientifically grounded, monetised impact quantifications for any company provides a defensible evidentiary foundation for regulatory disclosures that corporate sustainability teams and asset managers increasingly need to support with auditable methodology. The ability to apply the model to EU Taxonomy eligibility assessment, SFDR PAI calculation and double materiality analysis within a single platform creates an integrated compliance tool rather than requiring separate data sources and methodologies for each regulatory framework.

 

Read more: MAAS Subsidiary Huazhi Future Forms 800VDC Green Energy Research Team

 

Platform Capabilities and Decision-Making Applications

 

The Upright platform enables users to combine the model's sustainability intelligence with their own reports, operational data and business context to benchmark against peers, assess risks, challenge existing assessments and explore the evidence behind impact claims. This integration of external scientific evidence with internal company data creates a richer analytical foundation than either source alone can provide, allowing users to contextualise the model's general impact quantifications with company-specific operational information that can materially affect sustainability risk and opportunity assessments. The scenario modelling capability referenced by Nieminen enables forward-looking analysis that goes beyond historical disclosure review to project how changes in business activities, regulatory environments or physical conditions might affect a company's sustainability impact profile and associated financial materiality.

The monetisation of sustainability impacts in financial terms is a particularly important capability for connecting ESG analysis to mainstream investment and corporate strategy decision-making, where environmental and social impacts need to be expressed in units that are comparable with financial performance metrics to inform capital allocation decisions. By quantifying the value of positive and negative externalities in monetary terms alongside conventional financial data, the Upright model enables a more integrated assessment of company value creation and destruction that encompasses impacts on natural capital, human capital and social systems that conventional financial analysis excludes. This integrated value perspective is increasingly relevant for mandatory double materiality assessments under CSRD, where companies must evaluate both the financial materiality of sustainability risks and their own impacts on society and environment.

 

Explore OneStop ESG Marketplace: ESG Software

 

Outlook for Science-Based ESG Data Platforms

 

The launch of Upright's vertically-trained sustainability AI model reflects a broader evolution in the ESG data market away from disclosure-based data aggregation toward evidence-based impact quantification that can provide more decision-useful information for companies and investors navigating complex regulatory requirements. Whether Upright can demonstrate sufficient accuracy, consistency and coverage to compete with established ESG data providers including MSCI, Sustainalytics and S&P Global Sustainable1 will depend on the quality of the scientific evidence underpinning the model, the breadth of company and sector coverage and the practical usability of the platform for compliance and investment applications. Sustained adoption by corporate sustainability teams and institutional investors would validate the science-based approach as a commercially viable alternative to disclosure-dependent ESG data methodologies.

The convergence of mandatory double materiality requirements under CSRD, growing investor demand for decision-useful sustainability intelligence beyond compliance reporting and the improving capability of domain-specific AI models creates conditions in which vertically-trained sustainability AI platforms can capture significant market share from general-purpose tools adapted to sustainability applications. The next phase of ESG data market development is increasingly likely to be defined by the quality of underlying evidence and the coherence of impact quantification methodology rather than by data coverage breadth alone, making Upright's science-first approach a potentially significant competitive differentiator if the model delivers the consistency and reliability that Ojala describes.

 

Source: Upright Project