The technology industry has long sold us a story of frictionless progress, where innovation arrives with the clean efficiency of a software update. This week’s news suggests a different, messier reality: the infrastructure of artificial intelligence is creating tangible costs—in energy, in privacy, in geopolitical tension—that are no longer abstract.
Consider the two-pronged crisis emerging from AI’s insatiable appetite for data centers. The surge in demand is driving up electricity costs for American manufacturers and simultaneously causing a global memory chip shortage that raises prices for consumer electronics 1. This is not a future hypothetical; it is a present squeeze on industrial competitiveness and household budgets. The same week, a major Telstra outage in Australia halted regional trains, disrupted payments, and affected emergency calls, sparking a regulatory investigation 10. While not directly AI-related, the fragility of digital infrastructure is a reminder that when systems fail, the consequences are immediate and physical.
The regulatory landscape is shifting to match this new reality. The U.S. Department of Commerce has approved a broad public rollout of OpenAI’s GPT-5.6 model series, ending weeks of restricted access under a new government oversight framework 2. Meanwhile, Apple lost its legal challenge against EU rules designating it as a “gatekeeper” under the Digital Markets Act 9, and Nintendo will end original Switch sales in Europe due to new EU regulations requiring user-replaceable batteries 7. These decisions, from different jurisdictions, share a common thread: governments are asserting control over technology’s physical and economic footprint, not just its software.
The privacy implications are equally stark. Meta launched its Muse Image AI generator, which allows users to tag any public Instagram account to incorporate that person’s likeness into AI-generated images without notifying them . China’s National Vulnerability Database warned that Anthropic’s Claude Code contains a “security backdoor” that can transmit sensitive user information . And U.S. lawmakers are investigating the growing adoption of Chinese AI models by American companies, while Beijing weighs restricting overseas access to its own advanced models . The question “who benefits?” becomes urgent when data flows across borders with little accountability.