AI's Hidden Trash Problem Is Bigger Than Anyone Thought

A new report from the Basel Action Network (BAN), a nonprofit focused on hazardous waste, warns that AI's garbage problem has been badly underestimated — and the numbers are stark, according to The Verge.

Previous studies counted discarded servers and GPUs and called it done. BAN counted everything: cooling systems, power distribution, backup generators, networking gear, and what it calls "AI Waste Contagion" — the phones, laptops, and other personal devices that get replaced earlier than they otherwise would because AI advances make older hardware obsolete faster. That broader scope changes the picture dramatically. Past studies missed roughly 87 percent of a data centre's physical infrastructure.

BAN's conclusion: between 2025 and 2050, AI-related equipment could generate 395 to 617 million metric tons of e-waste. By 2050, global e-waste overall could triple to 211 million metric tons a year, with AI accounting for 15 to 20 percent of that.

The disposal problem is as serious as the volume. Less than a quarter of the world's e-waste is formally recycled. Much of the rest ends up burned or buried in informal recycling operations, exposing workers — and millions of children who live nearby — to lead, chromium, and other toxic materials. The US, home to more data centres than any other country, has not signed the Basel Convention, the international treaty meant to limit cross-border shipment of hazardous waste.

There is no immediate action for an individual here. What to watch: whether major AI companies begin publishing plans for hardware end-of-life, and whether the US moves toward Basel ratification. Those are the two signals that would suggest the industry is taking this seriously.

Google Home MCP is a new integration that lets third-party AI agents — tools like Claude, rather than Google's own assistant — read your smart home data and control your devices directly, according to The Verge.

MCP stands for Model Context Protocol (a standardised way for AI agents to connect to external systems, like a universal plug socket for software). The integration means you could ask Claude to review your camera footage from the last hour, track how many laundry cycles ran last week, or build a custom control dashboard — in plain English, through a chat window.

It is aimed at people who already pay for Google Home Premium Advanced ($20 a month or $200 a year) and are comfortable setting up a Google Cloud project to get it running. That is not a weekend-click-through setup.

Google says safety limits are in place — an agent cannot unlock your front door — but the company's own product manager notes that connecting an agent to your home "can result in unexpected or even undesired behavior."

Honest read: Worth watching if you run a Google Home setup and enjoy tinkering, but the setup friction and the $20/month requirement make this a tool for enthusiasts rather than most users — and the security implications of giving a third-party AI agent access to your locks and cameras deserve more scrutiny than a blog post caveat.

What it is: A formal process OpenAI published for tracking, investigating, and disclosing cases where its AI models behave in ways they were not intended to, according to OpenAI.

What's new: Previously, OpenAI disclosed misalignment cases sporadically — waiting until it had several incidents to bundle together, or burying findings in technical documents released alongside new models. The new framework commits to publishing reports faster, even before the behaviour is fully understood or fixed.

How it works: When a model behaves unexpectedly — acting without authorisation, attempting to coordinate with other AI systems, evading oversight, or behaving in ways that contradict published safety assessments — the framework triggers a formal investigation and disclosure report. Critically, disclosure is supposed to happen even when the significance is uncertain. The framework also covers model behaviour throughout the full lifecycle: training, testing, and live deployment. Six initial reports covering the last six months have been published alongside the framework itself.

What it can't do: OpenAI acknowledges there are no industry-wide standards for this kind of disclosure yet — this framework is a unilateral commitment, not an agreed benchmark. Some of the disclosed cases may turn out to be statistical noise rather than meaningful patterns. Verification is left to outside researchers examining the published reports.

Who should care: Anyone following AI safety debates, researchers studying model behaviour, and policymakers — this is one of the more concrete transparency commitments a frontier AI lab has made in public.

Read it: openai.com

Anthropic has launched Claude Docs and Claude Slides in beta — tools for creating shareable documents and presentations directly inside a Claude conversation, as reported by The Verge.

You describe what you want, Claude builds it, and you can edit manually or ask Claude to revise in plain language. Finished files live at a shareable link and can be opened for collaborative editing by other users, with comments working similarly to Google Docs. Alongside this, Anthropic has merged its separate Cowork and regular chat interfaces into a single "one Claude" experience, so these tools are available from any conversation without switching modes.

If you already use Claude Pro or Max (web, desktop, or mobile), this is rolling out now and requires nothing to switch on. Team and Free users get access later.

To start: open a Claude chat and ask it to draft a document or presentation on any topic.

Worth knowing: Claude Docs and Slides are in beta, and both your inputs and outputs are subject to Anthropic's standard data practices — check those before pasting anything confidential into a generated document.

The European Commission has proposed the EU KIDS Act, which would ban children under 15 from independently creating accounts on social networks and video platforms across all EU member states, according to the European Commission.

The goal is to replace the current patchwork of different national age rules with a single EU-wide standard. For ordinary users and parents, the practical effect would be that platforms must verify age before allowing account creation — something that currently happens inconsistently, if at all. This is a proposal, not yet law. It must pass through the European Parliament and Council before taking effect.