SAP at Climate Week: AI's Future Depends on Energy, Data and Trust

Sustainability returns, on different terms

New York Climate Week ran alongside the UN General Assembly with more than 1,500 events across the city and was hailed as the biggest Climate Week yet, SAP says. The writer describes a week focused on delivery and implementation rather than ambition, shaped by an energy price shock, a summer of climate disasters and a growing argument over the resources AI consumes.

On energy, UN Climate Chief Simon Stiell put the cost of energy shortages linked to the Iran war at more than $100 billion in higher fuel prices for American consumers alone. The UK's Ed Miliband urged governments to treat climate and nature loss as a matter of national security. The General Assembly also approved its first declaration on sea-level rise, and the UN has acknowledged that overshooting 1.5°C is now inevitable.

For companies, an MSCI analysis estimates corporate business interruption losses from climate hazards at more than $1.1 trillion a year, most of which never appears as a line on a balance sheet. SAP argues that many firms report climate risk but far fewer decide with it. To address that, SAP and Jupiter Intelligence are connecting Jupiter's physical climate risk intelligence with SAP's enterprise data and reporting foundation.

The AI debate: permission to operate

AI dominated the week, and views were divided. Stiell warned that AI leaders are "on thin ice" when it comes to their license to operate, while other speakers from business and academia argued AI is too important to give up and will help tackle climate change.

SAP's view is that the deciding issue is not capability but whether the people who host AI, power it and work alongside it keep agreeing to it. Every model runs on land, electricity, water and cooling, and every data center needs a willing community.

Governments are starting to set expectations. Türkiye's COP31 presidency launched the Antalya Pledge on AI. The COP31 President-Designate said companies are expected to be transparent about their energy use and to power their operations with clean energy. A draft is due before the summit, which will have a dedicated day on technology and AI.

SAP says it has run its own offices and data centers on 100% renewable electricity since 2014, considers environmental impact at the design stage of its AI architecture, and has started sharing the footprint of its cloud services with customers. A beta view for selected products is now in SAP for Me.

Governance and connected data

SAP also pointed to the upside of AI. Working from connected data, AI can weigh cost, carbon, availability and regulation in a single decision. "AI can weigh cost, carbon, and risk together, but only if sustainability is in the data it works from. Leave it out, and AI will optimize it away," said Sophia Mendelsohn, chief sustainability and commercial officer at SAP.

On governance, SAP says that in 2025 it reviewed 860 AI use cases against its AI ethics principles and escalated 44% as high-risk for senior review. Where no ethically acceptable trade-off could be found, the project did not go ahead in its original form.

The main obstacle to implementation, according to the article, is disconnected data. When sustainability information sits apart from procurement, manufacturing and finance, even good AI can produce confident wrong answers, such as recommending a lower-carbon supplier that is not approved for the market in question. Industrial manufacturer HARTING connected its sustainability data to its existing systems, moving product carbon footprint calculations from manual work to AI-assisted mapping across its material portfolio.

What it means: the road to Antalya

New York Climate Week was the last major business gathering before COP31 in Antalya in November. The COP31 presidency used it to launch a global pledge to raise electricity's share of final energy consumption to 35% by 2035, alongside the new pledge on how AI is powered and measured.

For companies, SAP says customer conversations have shifted toward what carbon costs, where supply chains are exposed and whether the numbers will hold up to an auditor. The company's position is that the tools, economics and business case are largely in place, and what remains is connecting them inside the processes where decisions are made. Expect growing pressure on AI providers and users to measure and disclose energy use.