Britain's foreign secretary accuses settlers of carrying out the "ethnic cleansing" of Palestinians, as Israel responds by shutting the UK consulate in East Jerusalem.
A recent multi‑year compute agreement that makes OpenAI an anchor customer of Firmus has pushed the firm’s contracted capacity in Australia past 900 megawatts. That single fact speaks to a broader truth: the rapid roll‑out of large AI systems has tangible consequences for electricity systems, communities and public policy. Australia must reckon with them now, before ad‑hoc capacity builds strain on grids and on public consent.
The technology race is not just about chips and models; it is about real world kilowatts. Making a major AI provider an anchor customer signals long‑term, concentrated demand — power delivered day and night to run servers, cooling systems and supporting infrastructure. Bringing contracted capacity across a firm to more than 900 MW in the region is not a footnote for planners, it is a material change in the scale of industrial electricity demand.
For everyday Australians that matters in practical ways. New large‑scale electricity demand can drive investment in generation, network upgrades and storage — potentially good for jobs and economic activity. But it can also exacerbate supply challenges, push up peak network costs for households and businesses, and complicate emission reduction plans if the additional load is met with fossil‑fuelled generation rather than renewables and storage.
Policy choices will determine the balance. If governments and planners treat this surge as merely a private matter for firms to sort out, there is a risk of hurried infrastructure being built in places least able to resist it, or of growth being fed by high‑emission power. Conversely, a planned approach can steer new investment towards off‑grid renewables, paired storage, or industrial clusters that create local jobs and share benefits with host communities.
The deal underlines the importance of transparent, forward‑looking energy planning. Governments and grid operators need better visibility of long‑term compute contracts so that network upgrades and generation capacity can be coordinated. Communities where large facilities are sited deserve clear information about local impacts and tangible community benefits — from training and jobs to investment in local services.
There is also a role for conditions and standards. Where anchor customers expect long‑term, round‑the‑clock supply, governments should insist on credible pathways to low‑emission power — whether through direct procurement of renewables, long‑duration storage, or verified clean energy imports. Planning and environmental approvals must treat these projects like any other large industrial customer: assess cumulative impacts, test grid resilience and set community expectations.
Australia has been a welcome host for technology investment. But the scale of modern AI compute means that hosting comes with responsibilities. The nation should get ahead of the curve: require better forecasts of large electricity contracts, align new demand with decarbonisation goals and make sure local communities see the benefits. If we fail to do that, we will find our electricity system rearranged not by democratic debate, but by commercial deals done elsewhere — and the consequences for prices, emissions and communities could be profound.
Community angle
As AI companies sign large, multi‑year compute deals, Australians face real impacts on local electricity supply, prices and jobs. The editorial calls for transparent planning, conditions linking new demand to clean power, and community protections so benefits are shared.
Australia is on the brink of a meaningful shift in online regulation. The government’s proposals to let social media users opt out of algorithm-recommended content — part of the so-called “My Feed, My Way” push — are a welcome step in confronting how invisible systems shape what we see, how we think and how our children grow up. But good intent alone will not protect the public; the law must be clear, enforceable and resourced to bite.
Why this matters to ordinary Australians is obvious. Algorithms power feeds on Facebook, Instagram, TikTok and other services, deciding what appears in front of us in seconds and nudging attention, habits and risk. The government’s brief lists real harms to be tackled: content that promotes eating disorders, hostile ideas about women, pornography, crime, dangerous stunts and material that risks serious mental harm for young people. These are not abstract policy problems — they are the kinds of content parents, schools and communities are trying to defend children and vulnerable people from today.
The proposal to give users the option to switch off algorithmic ranking is sensible in principle. Choice recognises that not everyone wants a perfectly personalised, engagement-maximising feed. Some users prefer chronological timelines or neutral streams; others want stronger controls to prevent addictive loops. Curbing algorithmic amplification can also blunt the commercial incentives that make extreme, sensational and harmful content more profitable for platforms to promote.
But a law that merely requires platforms to add an “opt out” toggle and then leaves the rest to corporate design will not protect young people or public discourse. Implementation details matter. What exactly counts as an “algorithmic” feed? How will platforms present the choice so that it is meaningful rather than buried in settings? Will the option be available on all devices, in all apps and to all account types? And crucially, how will regulators verify compliance?
The government has reportedly signalled tough penalties for non‑compliance — figures that have been reported up to $100 million. That is the right instinct: penalties must be large enough to matter to multibillion‑dollar companies. But fines alone are not a silver bullet. Regulators need the power to audit systems, demand technical documentation and test user experiences. They must be able to require changes to user interfaces that currently nudge people towards the algorithmic default.
There are other risks to manage. Heavy‑handed rules that treat all platforms the same risk squeezing out smaller services and entrenching incumbents. The law should be proportionate and technologically neutral, focused on outcomes — transparency, meaningful user choice and demonstrable protections for children and vulnerable people — rather than prescribing a single technical fix.
This is also a moment for broader civic conversation. Schools, parents and community groups must be supported to teach digital literacy and resilience at the same time governments legislate. Opt‑out controls are part of a toolkit, not the whole solution.
Australia should be applauded for trying to take the lead. If this legislation is to be more than theatre, lawmakers must harden vague commitments into specific, testable rules; give regulators teeth to audit and enforce; and ensure the protections are practical and visible to everyday users. Otherwise the next generation will simply be handed different looking feeds that behave in the same harmful ways. That outcome would be unacceptable — and avoidable.
Community angle
Ordinary Australians — parents, teachers and young people — need practical protections from harmful online content. The editorial urges laws that give meaningful choice, strong enforcement and clear standards so that opt‑out controls are real, not cosmetic.
Former Chancellor Angela Merkel called the results in Saxony-Anhalt "simply shocking," and warned of the dangers of feelings trumping facts. The latest PISA education test scores also have Germany reeling. DW has more.
HMM has signed a 25-year shipping contract with Vale worth approximately KRW 4. 7 trillion (US$3. 4 billion) . Under the agreement, HMM will transport iron ore for the Brazilian mining company from 2030. The latest deal is HMM’s third major long-term contract with Vale.
