Artificial Intelligence

• Misinformation have always existed
• Code algorithms have hardcoded biaised since the Internet was created
• AI is learning from all of that
• You can't trust AI
• … but you can't trust the web

Ethical Principles for ML

Ethical Principles for Web Machine Learning
W3C Group Draft Note, 8 January 2024

• Documents ethical issues associated with using Machine Learning on the Web
• general consideration of harms, risks and mitigations relevant to Web ML
• Accuracy: deviation from a true value can affect life, including credit scoring, loan approval
• Bias: systematic deviation can disproportionately affect individuals or groups
• Privacy: operating without a user’s knowledge/consent, scraping personal information to train models

Fast-pacing reality

• LLMs and agent protocols are evolving fast and competing
• We're still climbing the hype curve
• Regulators are trying not to lag
• We're all learning (see also AI slop, AI Darwin Awards)

So, why AI?

As users and developers, our modern world is more complicated.

• Convenience: we're lazy
• Ease of use: we want to get stuff done
• Cost: we're cheap

Why AI and the web?

• The web contains a LOT of information and services, including new information and services every day
• AI needs to leverage the web to be useful
• Keep in mind that a majority portion of the Web is not accessible through search engines

Web Consumers

• Humans: "This is for everyone" (Tim Berners-Lee, 12 August 2012)
• Humans, with assistants: Browser extensions, Translators, AI agents, etc.
• Tools: aggregators, crawlers (search engines, AIs), etc.

AI and Publishers

• Search crawls vs AI crawls have different relationships/interests
• AI opt-out becoming more prevalent, reducing the availability of data to AI:
  • lack of attribution for data sources
  • lack of an incentive for publishers
  • lack of any means of monitoring compliance

IAB AI-CONTROL Workshop Report,
IETF Internet Architecture Board, 6 September 2025

AI and user applications

AIs need to:

• Help the applications: image (face recognition), audio (speech recognition), text (translation), sensors and data (GPS and semantic segmentation)
• Help the user: Get me a pizza, Go to a music concert and tell my friends about it, …

What do we need?

1. Connect the applications to AIs
2. Connect the AIs with ALL applications (native, web, miniapps, etc.): calendars, social platforms, wallets, online shops, etc.
3. Connect the user to the AIs

Connecting applications with AIs

Applications need to leverage LLMS, and run new LLMs

• High-performance API to load and run an LLM
• APIs to leverage Client side, preinstalled, LLMs

Run your own LLM: Neural Network API

Web Neural Network API

• Hardware-agnostic abstraction layer for NN inference
• Use cases: Person Detection, Style Transfer, Image Captioning, Detecting fake video
• Sync/Async build and execution, device selection (cpu/gpu), power preference
• Operands: sigmoid, softmax, slice, gru, hardSwish, squeeze, etc.

Leverage client LLMs: built-in AI APIs

• Leverage existing LLMS (edge computing)
• LLMs: preinstalled or accessible as a service
• Access using API: translator, summarizer, writer proofreader, prompt (for web extensions)

Connecting AIs with applications

AIs needs to understand the application intent, and interact with the application:

enable agentic browsers: conversational, task automation, context awareness, integrated AI tooling

• using AI:

  markdown, screenshots, and code analysis (eg Overlays)

• using developer hints: AI agents struggle to navigate existing human-first interfaces

  declarative (metadata) and programmatic (APIs)

• Going beyond: generative UI: different layouts and components to different users

Using developer hints

• Native applications: see agent to agent protocols (MPC, etc.), accessibility APIs, etc. For Web browsers, see WebDriver.
• Web applications:
• declarative:
  • Accessible Rich Internet Applications (WAI-ARIA)
  • see LLM & Linked Data (AC Meeting 2025)
• API: see WebMCP (app-controlled UI), AOM
• Server applications: see agent to agent protocols (MPC, etc.)

Connecting the user to the AIs

• Agentic applications: Chat, Suggestion
• human-in-the-loop: Event-driven communication and bidirectional interaction
• frontend integration: AG-UI, MCP-UI, NLWeb
• Client-side generative UI: meeting the needs of the user

Agentic Web?

A World Wide Web that connects and empowers humanity, enables human-AI interactions, based on ethical, privacy, and security principles.

Thank you

W3C brings together global stakeholders to develop open standards that enable a World Wide Web that connects and empowers humanity.

See also September 16 TAG AI discussion