Artificial intelligence (AI) has enormous potential to improve business practices, but the technology is limited in its ability to solve complex social sustainability challenges and should not be relied on to produce sustainability reports, experts said at an event in Hong Kong.
To continue reading, subscribe to Eco‑Business.
There's something for everyone. We offer a range of subscription plans.
- Access our stories and receive our Insights Weekly newsletter with the free EB Member plan.
- Unlock unlimited access to our content and archive with EB Circle.
- Publish your content with EB Premium.
Speaking at the Unlocking capital for sustainability conference on Thursday, JP Stevenson, head of market development for supply chain assurance company LRQA, said that AI cannot tackle “hard” sustainability problems such as detecting forced or child labour.
“AI is primarily geared around information economies and consumer economies. That is how the technology is built. But the reality of the Bangladeshi factory is very different,” he said, referring to AI’s limitations in addressing social challenges in the non-digital world.
The potential of AI depends on its ability to digitalise supply chains – or else risk going the way of blockchain, which despite being hailed as a panacea for transparency and efficiency 10 years ago, has rarely scaled, Stevenson said.
Big technology companies have been heavily pushing AI as a sustainability problem-solver in recent years. Google marketing points out that the technology is being deployed to identify urban heat islands, reduce emissions at traffic lights and improve traffic flow in cities, among other uses.
Meanwhile, the impact of surging AI consumption on the climate and water resources is becoming more evident. Jinxi Chen, Hong Kong director at sustainability consultancy Pollination, noted that the high level of water used by data centres that power AI poses a business risk for companies in water-scarce parts of Asia.
A typical hyper-scaler data centre consumes water equivalent to the volume used by 30,000 to 50,000 people every day, and there are 8,000 data centres around the world – a number that is growing to meet rising energy demand, she noted.
Big tech firms have been among the biggest procurers of clean energy but 60 per cent of the energy consumed by data centres still comes from fossil fuels, Chen said. As data centres require constant power supply, they make the transition away from fossil fuels more difficult, she said.
“AI right now is so geared towards consumer tech, and we’re not necessarily applying the best tools for the job,” she said, adding that there’s a big gap in education for how people can use AI tools responsibly.
AI, reporting and greenwashing risk
JP Stevenson (left) of LRQA and MioTech’s Jason Tu at Unlocking capital for sustainability. Stevenson cautioned against relying on AI to produce sustainability reports. Image: Eco-Business
AI is increasingly used to process data used in corporate sustainability reports, which are growing in frequency as more countries mandate sustainability reporting for listed and large companies.
Jason Tu, founder and CEO of MioTech, a Hong Kong-based sustainability data and software startup, said his company will soon produce a tool that can replace the work done by investor relations teams and editors in weaving sustainability data together to produce a report. The tool can even mimic the tone of the firm’s corporate voice, he said.
However, Stevenson cautioned against relying on AI to produce sustainability reports, because the technology could amplify inaccuracies in the data, increasing greenwashing risk.
“It is wonderful that sustainability teams will no longer need to focus on reporting [if they use AI]. But one of the problems with that is that sustainability reports are often factually incorrect,” he said.
A global study of finance, technology, and sustainability professionals by GlobeScan in 2024 found that fewer than three in 10 (27 per cent) say they have access to high-quality sustainability data.
“How do we mitigate the risk associated with putting that [sustainability] information in a black box where you’re removing a manual, human-based review process?” he said. “If anything, it seems like we’re moving further away from accurate reporting.”
Before AI is used at scale for sustainability reporting, regulation must drive standardisation of supply chain data, which is currently lacking, he said.
