@llibertatuk
Design for humans. End to end, front to back, user-centred digital services with real purpose and impact for the public, private and third sectors globally. Led by former GOV.UK Head of Content Padma Gillen. Accessible by design. Posts by Katrina (she/her)
An appointment is an activity. A healthier life is a result.
If we want to move the needle on climate or health, we have to stop measuring what we spent and start measuring the difference we made.
Read why the project model is broken: transform.wales/blog/2026/fr...
#Wales #PublicSector (2/2)
The lesson here?
AI governance matters more than ever.
Think beyond hallucinations - even malicious actors.
No one's in control.
AI agents will fail, dangerously, even by acting benignly.
Embedded governance in infrastructure is crucial.
And that needs the whole UX team.
arxiv.org/abs/2602.20021
Where do content design and content strategy fit into an AI-driven digital service lifecycle?
- data curation and preparation
- training models and iterating
- creating checkpoints and validation steps
- optimising for ethical, trustworthy, inclusive interactions
Everywhere, in other words.
“Insisting on technical precision when it matters can help us resist the easy idea that AI is a helper, ready to assist. Instead we must keep our focus on the uncomfortable thought that these systems partly and covertly structure what we say, think and learn.”
🎵 You see, I’ve been through the desert on a
This. Digital services are doomed.
Sometimes the best ideas arise from moments when you feel simply 'unoccupied.' How could you structure your day to include intentional moments of boredom?
buff.ly/mtLVxh8
#amWriting #writingAboutWriting #creativity
great piece by @marietjeschaake.bsky.social with this pointed conclusion
Hey.
Tell people when they do good job.
When you like the stuff they make.
When you like the characters they’ve created.
The stories they’ve told.
Nobody hears it as often as you think bc everyone thinks everyone hears it all the time.
Trust. They don’t.
Tell. Them.
We keep framing AI like a person, but that changes how we trust it, fear it, and hand power over to it.
Watch now on YouTube: www.youtube.com/watch?v=47CI...
What's AI really for? Doesn't Hannah Fry's AI Confedential TV series show that designing for humans means questioning the AI user need? Being more productive? Or adding value to life? Not just functional, but emotional. All this and the real magic of transformation. www.linkedin.com/feed/update/...
Noted in the great New Scientist interview with Hannah Fry about her new series AI Confidential:
* humans are vulnerable
* it shouldn't have to be on the individual to protect themselves
* AI needs governance
So, collective action and start with user needs, right?
www.youtube.com/watch?v=iitq...
Starting with user needs - which is all about identifying what's actually needed, where, when, how, and if anything's needed at all, would definitely be great.
Reminder to always factor in risk to your business or organisation when you are going to be dependent on any tool, and how you'll manage those risks (or find alternatives). The strategy isn't delivery if you're building on foundations that might crumble.
A very interesting ad. Trust is everything for humans, especially online.
Before writing new content, ask:
can we improve what we have?
That's even more true with agentic AI.
Fix what's outdated, irrelevant, or just bad.
Discover the questions people ask online and provide answers quickly, clearly and authentically.
More does not equal better.
Better is better.
Ethical practice in research isn’t just about forms or approvals. 🧪
It’s about judgement, relationships, and care over time, especially when working outside formal systems.
I’ve pulled together a knowledge hub of practical protocols and resources here:
learningforsustainability.net/human-ethics/
Hah. I was reading a breathless article about how AI Agents were using a social media site created just for them yesterday and you probably felt the planet shift briefly out of rotation from the amount of deep sighs I made.
Well, surprise surprise it's still Mechanical Turk time.
I'm glad our hooting and hollering is making a dent.
AI is a tool, not magic fairy dust. As such, we have to be real straightforward about what it can and can't do... And what it shouldn't do. We are the ones wielding it, and we are responsible for what happens next.
please use alt text when posting images! especially for things like this, everyone needs to have access to this information—whether they’re a screenreader user or in a place with bad internet that will load the alt text but not the images.
ok but hear me out, what if we have an actual rebel alliance of fully non-genAI organizations that care about people and user needs first
Absolutely, all points to design for humans, steered by humans. As Max Tegmark says, the more we rely on technology the more it needs to be robust and secure, and that means human abilities to identify needs and proactively verify, validate, secure and control (monitoring and managing AI behaviour).
If nothing else* this announcement shows how rapidly events are overtaking policy these days.
*Oh mate
Great conversation. Would add starting with good UX. What are the business and user needs? Is AI solving a real problem for users? Embed trust, clarity and safety in strategy, including governance, and delivery - as one small example, providing users with pathways to review and verify AI outputs.
AI's acceleration makes Responsible AI (protecting human safety and security) tricky.
Here's some ways content design can help:
- putting user needs first
- auditing and iterating AI outputs
- designing words, actions and visual cues so that users are in control and can even verify AI outputs
Want to be "AI-ready"?
Stop chasing the latest AI tools. Focus on core UX skills.
Companies are still run by humans.
You need to solve human problems.
Hone:
- identifying user needs
- developing usable insights
- strategic storytelling
- joined up integration of new tech with legacy systems
Can Generative AI improve public services? Discover the answer at our FREE webinar where our panel of experts will discuss the implications of using AI in the public sector. With @emercoleman.bsky.social @richardpope.org @elenasimperl.bsky.social & Andy Dudfield from Full Fact. buff.ly/XtumBxC
A graphic with arrows representing positive and negative, and the text: How are you feeling about AI? So far, respondents to our 2026 AI survey are feeling marginally less positive about the increased use of AI in their daily professional lives, compared to last year. Do you agree or disagree? Please take ten minutes to complete the anonymous survey, and make your opinions and needs heard! All responses anonymous. Closes in February 2026
👀 Our AI survey results so far show that respondents are feeling slightly less positive about the increased use of AI in their daily professional lives, compared to last year.
Do you agree or disagree? Please take ten minutes to share your opinions & needs via our anonymous survey 👉 buff.ly/fsvJaZr
I've revived an old conference talk from 2023 which started life as a very short blog post. It's about how overwhelming it can be to work in content - and why doing less might actually be the answer. lapope.com/2026/01/16/o...
Good content online has a purpose.
Good content answers people's questions - quickly and easily.
What questions does your content answer?
If you don't know, users (including AI) won't either.
Make every piece of content work for everyone.
Discover what people are asking, and meet those needs.
