Daily Briefing — June 23, 2026

Australia has about 160 data centres operating right now, with another 90 proposed. The investment pipeline is estimated at $155 billion over the next decade. One planned "hyperscale" facility in Sydney's outer west would span 52 hectares, include six four-storey buildings, 936 cooling units, and 852 diesel backup generators. That is what the AI infrastructure buildout looks like on the ground: concrete, diesel, and water at industrial scale.

These facilities consume enormous amounts of electricity and water, generate noise and air pollution, and displace farmland and open space. The operational jobs they create are minimal compared to their footprint. Communities near proposed sites are pushing back, and the tension between national AI ambitions and local environmental costs is turning into one of the defining infrastructure fights of this decade.

Every model you use, every cloud query you run, has a resource cost that lands somewhere specific. Who bears that cost matters as much as the technology itself.

A Guardian investigation found brands using AI-generated influencers to promote products on social media. The content is designed to look like genuine customer experiences, with no indication that the people featured are not real. Some creators involved in making these campaigns are being asked to sign NDAs so they cannot talk about the work. Most markets have no rules requiring brands to disclose when ad content was created using AI.

Same news cycle, same playbook: Polymarket paid dozens of creators to film themselves making fake bets on dummy copies of its website, then post the videos as if they were real. Manufactured authenticity at scale, and the tools just got cheaper and harder to spot.

The EU's Artificial Intelligence Act will start requiring AI-generated content to be labelled in August. Most other markets have nothing similar. What is technically possible to fake and what is legally required to disclose are moving in opposite directions, and consumers are stuck in between.

Google's DeepMind lab is partnering with A24, the film studio behind Everything Everywhere All at Once and Moonlight, to develop AI-powered movie production tools. Google is investing around $75 million into A24 as part of the deal — the first time Google has taken a stake in a film studio. The stated goal is to build tools that help filmmakers "expand their storytelling possibilities," with an initial focus on production and distribution workflows.

A24 is the most filmmaker-friendly studio in Hollywood, known for giving directors creative control. Google is framing this as AI serving filmmakers, with tools shaped by the people who will use them. Whether that framing survives contact with actual products remains to be seen, but the choice of partner tells you Google knows the optics matter.

Every major AI lab is racing to embed its models in creative workflows — writing, music, design, now film. Google betting $75 million on A24's credibility with artists says that technical capability alone won't win this market.