Eight DAM Decisions for Drupal Teams
Your next website visitor might not be human
For most of the web’s history, we have designed digital experiences around a simple assumption: a person will visit our website. That person might arrive through a search engine, follow a campaign link, scan a QR code, or maybe even type the URL into their browser.
AI is changing that... dramatically and rapidly!
People are now asking AI assistants to research products, compare services, explain policies, recommend suppliers and complete tasks on their behalf. Sometimes, they might not even consciously choose AI and are simply guided by seemingly familiar tools like Google 'AI Overviews'. Either way, instead of visiting ten websites, a customer may ask one assistant to gather the relevant information and present a recommendation.
In the near future, that AI assistant could be doing more than reading a web page: checking product availability, requesting information, preparing an application, arranging an appointment or even completing a transaction.
Your next website visitor may not be a person at all, but an AI agent acting on their behalf, which raises a serious question:
Can AI systems understand our organisation, trust our information and interact with our services safely?
Getting to know your new audienceTo be useful, AI assistants need to find the right information, understand its meaning and decide whether it is current and trustworthy.
A prospective student asking an assistant to compare courses across several universities, a buyer requesting a shortlist of products that meet detailed technical, ethical and budget requirements - both are now part of your website’s audience.
While human visitors use navigation, page layouts, graphic cues and calls to action, AI systems depend more heavily on structured information, descriptive metadata, clear relationships and reliable access to data.
The closest thing I have to an answer
From Headless CMS to AI Harness: What I Took to Decoupled Days
Article by: Martin Anderson-Clutz. Originally posted on the Acquia blog.
Drupal turns decoupled architecture into a governed AI harness, combining live visual editing with agent-ready content schemas.
Back in March, at EvolveDigital in Toronto, I ran into Preston So. He mentioned that the team behind Decoupled Days was looking for speakers, and that this year the event would be in Montréal. I was interested right away. Drupal Canvas is the most compelling answer I have seen to a problem that has followed decoupled architectures for years, and I wanted that message to reach beyond the Drupal faithful — out to the practitioners who live and breathe headless every day.
The talk I ended up giving was not really about a content management system at all. It was about how Drupal has quietly become something else: a governed harness for artificial intelligence. Here is the argument I made, the demo that seemed to land hardest with the room, and why I think 2026 is the year the trade-offs of going headless finally stop being trade-offs.
Drupal Security Team Marks 16 Contributed Projects Unsupported in Ten Weeks
What Becomes Valuable When Code Gets Cheap?
This week's release of Drupal AI Context Beta 4 added taxonomy-based context selection, priority controls, workflow states, and token-budget visibility. Kristen Pol, primary maintainer of Context Control Center, said the project is moving towards its first release candidate, while Drupal.org release notes list 70 credited issues in the beta. The additions put more structure around deciding what information reaches an AI agent before it acts.
The release arrived as Drupal founder Dries Buytaert was asking a broader question about what happens when AI makes code cheaper to produce. In a 17 August 2026 blog post, Dries argued that as application functionality becomes easier to recreate, commercial value may shift towards dependable operation: deployment, security, scaling, monitoring, and reliable service over time. His argument is about software economics rather than AI context, but it points to the same larger change: implementation is no longer the only scarce input.
The DropTimes' 18 August interview with Kristen showed where that shift meets Drupal AI. Context Control Center can use moderation, revisions, permissions, scheduling, and scopes to govern the knowledge supplied to a model, but those controls cannot guarantee how the model will respond. Requirements that cannot tolerate probabilistic behaviour still need deterministic rules, validation, access controls, or human approval.
Umbraco on DDEV: .NET, SQL Server, and the generic project type

DDEV's generic project type will run anything that brings its own web server, and people already use it for Node and other stacks. Umbraco is a less-travelled case: .NET, Kestrel, and SQL Server. I wanted to know how much of the DDEV experience still holds up there. Clone a repository, run one command, have a working site and database a few minutes later.
Most of it holds up. The result is umbraco-clean-ddev, which runs Umbraco on .NET 10 with SQL Server and Adminer, and gets a new developer going with a single ddev start. What took the time was a handful of places where DDEV expected something I had not worked out yet.
I wanted four things out of this, and nothing more ambitious than that.
- One command for someone joining the project
- Nothing installed on my own machine: no SQL Server, no .NET SDK, none of the project's other dependencies.
- Projects that stay out of each other's way.
- A way to take a backup from Umbraco Cloud and debug against it locally.
DDEV already does all of this elsewhere. The Umbraco Cloud backup turned out to be the least transferable: ddev import-db and ddev export-db work on the database container DDEV manages, and this project omits that container entirely. Anything equivalent I would have to write myself.
Keeping the dependencies off my machine also makes the project easier to hand to someone who has never worked with .NET. The SDK and SQL Server both live in containers, so trying Umbraco does not start with installing either of them locally. Clone the repository, run one command, and the CMS is there to look at.
State of Drupal 7 in 2026: How Many Sites Are Still Running It?
The 10 Drupal SEO Modules That Matter in 2026 — and How to Configure Each One Properly
Drupal to Headless: The Node IDs That Break the Move
When you move a Drupal site to a headless CMS, the thing most likely to break silently is not your content. It is the invisible wiring between pieces of it. Drupal stores relationships as numeric IDs (node 4127, term 88, media 903), and almost every headless platform mints brand-new IDs the moment you import. So every taxonomy tag, every embedded image, every "related articles" block that pointed at an old number now points at nothing, or worse, at whatever content happened to inherit that number. The pages still render. The links inside them just quietly go nowhere. Here is the scene that keeps happening. The migration "succeeds." Every article is present, the word counts match, everyone high-fives. Then two weeks later someone notices the related-content sidebar is empty on 8,000 pages, half the article hero images resolve to a 404, and the tag pages that used to rank now list either…
From Headless CMS to AI Harness: What I Took to Decoupled Days
Talking Drupal #566 - DrupalEasy: Responsible Drupal AI
Today we are talking about Drupal, AI, and learning to use it responsibly with guest Mike Anello. We'll also cover Entity Mesh as our module of the week.
For show notes visit: https://www.talkingDrupal.com/566
I’m Joining Acquia!
A conversation in Cartagena
Why and how to migrate your DotNetNuke (DNN) site to Drupal
Submissions are closed. Nominations coming soon!
The International Splash Awards 2026 have reached a new milestone, with 40% more submissions compared to last year.
Image
A huge thank you to everyone who submitted a project and helped make this year’s edition even bigger.
The jury is now reviewing the submissions, with nominations set to be announced in early September.
We look forward to celebrating the projects and teams behind them during DrupalCon Rotterdam 2026.
Stay tuned. The nominations are coming soon.
DrupalCon Rotterdam 2026 T-Shirt Design Contest
Thank you to everyone who submitted a design and shared their creativity with the community, and to everyone who took part in the voting. The five finalist designs showed the imagination and community spirit that make DrupalCon so special.
Meet the winning design Image
The Drupal community has spoken!
After an open public vote, Juliane Vöske’s design has been selected as the official DrupalCon Rotterdam 2026 T-shirt.
Her winning design will be worn by attendees from across the global Drupal community. This is the T-shirt that will bring us together in Rotterdam.
See you there!
TDT August Town Hall Tracks Editorial Shift and Wider Community Outreach
Drupal code query MCP Server
A few months ago I put an api online to search and query the whole Drupal code ecosystem. With the release of some more capable local AI models I thought it could be good to release a MCP server with that same data optimized for LLM use. Same data, different packaging. The address for the MCP server is:
Add the server to your setup of choice and you'll be able to ask things like:
theodore August 20, 2026