Waste Management Business
AI panel at ISWA 2026: 'fog concealing the truth' and lazy research
ISWA 2026 panel warns LLMs are eroding trust in waste science: 'AI is creating a fog concealing the truth,' CIWM's Dan Cooke told delegates.

Waypoints
Panel 'Publishing Waste and Circularity Science in the Age of AI' held on day 2 of ISWA World Congress 2026
CIWM's Dan Cooke told delegates: 'AI is creating a fog that is concealing the truth'
Maya Zaatar has worked as a waste practitioner in over 25 countries in the Global South
Prof Mario Grosso of Politecnico di Milano warned of LLM sycophancy in scientific use
Panel called for a transparent, reliable source of truth for scientific papers and information
Day 2 of the ISWA World Congress 2026 put a hard question on the table: if large language models are now the default route to scientific information, who in the resources and waste sector is still verifying what comes out the other end?
The panel, titled 'Publishing Waste and Circularity Science in the Age of AI', was introduced by Dr Costas Velis, who framed the problem for delegates: AI is disrupting not just research outputs but the policy machinery that depends on them. LLMs such as Claude and ChatGPT, he argued, are changing how everyone perceives information and are actively eroding trust between the public and the scientific community.
What does AI do to scientific publishing?
Prof Mario Grosso of Politecnico di Milano added a sharper technical concern: LLM sycophancy. These models tend to agree with the human they are interacting with, which is precisely the wrong behaviour for a tool meant to challenge assumptions in peer review or evidence gathering.
Velis put the panel's central question plainly: how can we assess the accuracy and wider truthfulness of 'scientific info', and what tools do we actually have to do it?
Dan Cooke of CIWM answered with the panel's most quotable line. "AI is creating a fog that is concealing the truth," he told delegates. Cooke's diagnosis went beyond the technology itself: he argued we are living in a post-truth world where facts carry less weight than appeals to emotion, and where audiences have become accustomed to short snippets and impatient for an immediate truth. That consumption habit, he suggested, is what makes the AI problem acute rather than merely new.
Mark Shayler, circular economy expert and host of ISWA 2026, turned the critique inward. His concern was that AI is making practitioners lazy, and that the temptation to offload research work onto the models is massive. He drew a line practitioners will recognise: using AI as an indexing system is one thing; using it as a research or opinion system is another thing altogether.
"We have knowledge and tools to interrogate information," Shayler said. "We're the problem, we've become idle and lazier as the tools are so powerful, but the power comes with responsibility."
Does AI belong in Global South waste systems?
Maya Zaatar of the Royal Borough of Greenwich brought the discussion down to operational ground. Having worked as a practitioner in more than 25 countries across the Global South, she made a distinction that cuts against the technology-first instinct: waste itself is the same everywhere, but the systems that manage it differ from country to country.
For countries at the beginning of their waste infrastructure development, measurement is not the most important factor, Zaatar explained. The priority is delivering basic, simple, low-cost tools to manage waste — not integrating AI technologies. Local contexts should inform decisions, she emphasised, and academic papers are not always relevant to them.
Her point carried a corrective for the sector's framing. Circularity is not a theoretical concept imported from the Global North; it is an on-the-ground reality in many Global South countries already. She cautioned against putting circularity into one big basket — it looks different in different geographies, and that diversity should be embraced rather than standardised away.
Where does the sector go from here?
The panel closed on next steps, and the answers were deliberately unglamorous. Cooke argued the answer lies within ourselves: practitioners already know how to verify sources, and those verified sources can serve as a trustworthy guide through the fog of new information.
Shayler advocated a strategy of slowing down. There is real benefit, he argued, in slowing down and asking better questions before reaching for a tool. He also called for establishing who is responsible for highlighting papers and information — a transparent, reliable source of truth would be a first step toward solving the publishing problem in an age of AI.
For waste and circularity professionals, the takeaway is a working rule rather than a slogan: AI can index, but it cannot be trusted to agree its way to the truth. The sector's next test is whether it builds the verification infrastructure Shayler described — before the fog thickens.
via i0.wp.com (Original)
More from Elena Vasquez
Show full bio
Senior reporter covering technology and product news in Circular economy and the environmental industry at Circular Wire.
546 articles