The New Front Door to Government
'How do I renew my passport?' 'What documents do I need for a business license?' 'Am I eligible for this benefit?' These questions are already being asked to AI assistants millions of times per day. The AI agent visits government websites, tries to extract the answer, and struggles through Drupal's template-heavy HTML output. The citizen gets a vague or incomplete answer. The government loses a service interaction.
GOV.UK Shows the Alternative
The UK's Government Digital Service built GOV.UK on a custom Rails application with API-first architecture and clean semantic HTML. It's the global benchmark. When an AI agent queries a GOV.UK page, it gets structured, parseable content — not CMS template noise. Singapore's government digital properties follow a similar model. Both invested in infrastructure that serves machines as well as humans.
Most governments didn't. They bought Drupal, configured it once, and haven't touched the architecture since. The result: public service information locked inside frameworks that the majority of web traffic — AI agents — can't efficiently process.
The Democratic Access Problem
If AI-mediated access to government services becomes the norm — and at 57.5% bot traffic, it already is — then government AI-readiness becomes a democratic access issue. Citizens who use AI assistants to navigate government services will get better outcomes from governments with machine-readable web infrastructure. Citizens in jurisdictions running legacy Drupal will get worse outcomes. Framework choice becomes a matter of civic equity.
