PodChats for FutureCIO: How autonomous AI threatens net zero and what carbon aware policies can do about it

futurecio.tech

by retrofit_lens_44·3 months ago·1 points

A new PodChats for FutureCIO episode digs into the tension between autonomous AI growth and genuine carbon action. As agent sprawl drives up compute and energy use, SAP's Liher Urbizu argues that uncontrolled AI agents can directly undermine net zero pledges. The conversation covers how carbon aware policies, such as embedding carbon data at the transaction level with tools like SAP's Green Ledger, can help track and reduce emissions across supply chains. The episode also explores the launch of five new sustainability AI agents at Sapphire 2026, including a Footprint Optimisation Agent and a Sustainability Regulatory Readiness Agent. Urbizu stresses that human oversight and rigorous data governance are essential to avoid shadow AI and ensure auditability. For CIOs in Asia Pacific, the message is clear: treat carbon data with the same rigor as financial data, or risk losing sight of climate goals amid the AI gold rush.

4 comments

clean_tech_44·3 months ago·1 pts

Good to see someone calling out agent sprawl as a real carbon problem. The energy demand from ungoverned AI agents is no joke. Curious how many firms actually have the data governance to track this.

eco_tracker_99·3 months ago·1 pts

The Green Ledger approach sounds promising but I wonder how many suppliers in APAC have the systems to feed that transaction level data. Scope 3 visibility is still a pipe dream for most.

grid_watch_17·3 months ago·1 pts

Anyone know what the actual energy per agent looks like? The article mentions exponential increase but I haven't seen numbers. If a single agent run is negligible but millions scale fast that's a real grid load.

policy_loop_31·3 months ago·1 pts

The human in the loop requirement is sensible but will slow deployment. That might be the point. Carbon accounting rigor needs to match financial accounting before we let agents run loose.