Links

From 2 August, EU regulation around labeling AI generated content goes into effect.

Providers of AI generated content can label it however they want, but to make to it easier to know if you did enough, the Commission facilitated the production of a Code of Practice, that includes icons that can be use.

The Code of Practice on Transparency of AI-generated Content:

provides an EU-wide recognised practical framework for signatories to demonstrate compliance with those obligations

(From: Code of Practice on Transparency of AI-generated Content | Shaping Europe’s digital future)

There's a new release of the Web Sustainability Guidelines (WSG))

Updates include:

• New Impact Measurement + Impact API.
• Lots of small tweaks following peer review: we had at least 2 people look at each success criterion, consolidated and removed some (eg UX went from 21 to 17, web development from 20 to 16).
• Improved tone of voice and readability throughout, new introduction and “web sustainability” definition.

(Full changelog)

The new definition:

An approach to designing, developing, and operating digital products and services that leads to a web that is as clean, efficient, open, honest, regenerative and resilient as possible.

From the updated intro:

Web Sustainability Guidelines (WSG) provide actionable recommendations to help digital teams make informed, sustainable decisions. In considering different aspects of work and the web, these guidelines address the planet, people, and prosperity (PPP) impacts of digital products and services. They are interdisciplinary and cover artificial intelligence and emerging web technologies.

Dr. John Bistline, Head of Science at Watershed:

We're releasing an open framework to give companies a defensible starting point and a path toward increasingly precise AI emissions measurement as data availability improves.

(From: An open framework for AI emissions measurement – Watershed)

LOL:

The EUIPO found that the word “open” would be understood by the relevant public as meaning freely accessible, while the combination with “AI” (artificial intelligence) would be interpreted as referring to products based on openly accessible artificial intelligence

(From: OpenAI loses trademark dispute at EU court - Europe)

I blogged on the W3C site about the meetup we organised when the AB was in Rotterdam: CSS animations, human rights, and big accessibility at the AB Web Meetup | 2026 | Blog | W3C

Ketan Joshi analysed Google's latest sustainability report, which details enormous increases in energy consumption, which is not good news given we're in a climate crisis.

Adding up the numbers, Joshi says the increase isn't due to people manually prompting Gemini, but likely associated with the fact Google's search engine shows AI results by default:

GenAI’s climate harm comes from the system design, such as Google’s AI overview triggering constantly whether users want it or not

(From: Google’s exponential path to climate-wrecking digital bloat – Ketan Joshi)

They could choose to try and have the reverse effect, by designing it like the Spice Girls would: triggering GenAI-abled functionality only when users really, really want it.

Jo Lindsay Walton argues for Assured Interventions (AI), things that actually can contribute to the climate transition reliably:

massive and delicious demand-side mitigation of meat methane, nationalise LDAR for fugitive emissions, stop deforestation with community rights and loss and damages funding, ban cryptocurrency

(From: Why we need “AI” for climate transition | by Jo Lindsay Walton) (sadly on Medium, but it's a great piece)

Three findings from this week's UN report on AI vs climate crisis:

🌿 Majority of AI energy use (80-90%) is not in training models, but in day-to-day use.

🌿 Generating an image is 1450⨉ more energy intensive than text, generating video much worse.

🌿 We need to monitor water and land footprint, as well as carbon.

(From: Rising Emissions, Depleting Water and Vanishing Land—UN Scientists: AI Is Threatening Natural Resources for Billions | United Nations University)

The WebKit team responded to WebMCP in standards-positions.

What stood out to me:

When a site's actions are hard for an agent to use, that is a gap in the page's own semantics, and the fix, in our opinion, is to close it in the platform's shared layers (HTML and ARIA), where the user, assistive technology, and agents all benefit. Describing those actions a second time as JavaScript tools would not deliver the reliability it promises: the agent still selects and interprets a tool from its natural-language name and description, which the specification itself concedes are ambiguous and unverifiable, with "no guarantee that a WebMCP tool's declared intent matches its actual behavior."

(From: WebMCP · Issue #670 · WebKit/standards-positions)

Another reason to disallow LLM contributions in open source projects would be that it gets in the way of the project nurturing “new, confident, trustworthy contributors“, says (even) Simon Willison:

It doesn't matter if the LLM helps you submit a perfect PR to Zig - the time the Zig team spends reviewing your work does nothing to help them add new, confident, trustworthy contributors to their overall project.

(From: The Zig project's rationale for their firm anti-AI contribution policy)

(via Cassandra)