datacentre

Major technology companies are rapidly expanding the use of guarantees to back debt financing for AI datacentres and chips, issuing up to $300 billion in commitments in less than a year, according to analysis shared by economist Adam Tooze. The commitments largely stay off the companies' balance sheets, raising questions about hidden leverage in the AI buildout and how much risk the industry is actually carrying.

A plan for a $1.2 billion datacentre in Sydney has been abandoned, with emerging regulations cited as the reason. The project had drawn controversy, and the decision highlights growing uncertainty for large datacentre developments in Australia as governments move to introduce new rules governing the sector.

Industry observers are asking when the global memory chip shortage will ease. The squeeze is driven by surging AI infrastructure demand, limited wafer capacity, and the priority given to HBM production, while DRAM and NAND investment remains constrained. New manufacturing capacity takes years to build, so analysts say relief is unlikely before major fabs now under construction come online.

Anthropic will not appear at an inquiry into AI and datacentres, amid fallout from a reported hack of OpenAI. The company behind the Claude chatbot had been invited to appear, and its chief executive is expected to decline. Anthropic is instead expected to attend a separate Australian government hearing on artificial intelligence next week.

Concern is growing in Alberta, Canada, that the rapid build-out of energy-hungry data centres, driven largely by artificial intelligence demand, will translate into higher electricity bills for households. The issue has become a prominent public talking point in the province, touching on debates over energy policy, infrastructure capacity, the role of the oil and gas sector, and the environmental and economic costs of attracting big tech facilities.

NVIDIA is again emphasising the 'rentability' of its GPUs, framing graphics cards as revenue-generating assets rather than consumer purchases. The framing has drawn scrutiny from tech commentators and gamers, who question whether the company is nudging customers toward treating hardware as an investment amid high prices and the growing AI datacentre market.

Lea Weimann has published an overview of legal and regulatory options available to citizens who want to influence policy on the ethical and environmental impacts of AI and datacentres. The piece argues against the push to embed AI in every application and invites public discussion on how regulation could address energy use, water consumption and community concerns around datacentre expansion.