The COVID-19 Inquiry has published its damning assessment of government decision-making during the pandemic. Buried in nearly a thousand pages is a finding that should embarrass everyone involved in government digital services: the UK government and Northern Ireland Executive initially failed to provide British Sign Language interpretation for press conferences or key guidance in alternative formats.

The Inquiry states this plainly: “These are not secondary considerations. Everyone should be able to understand the action their government is asking them to take.”

I worked in the Cabinet Office during COVID-19. I watched this unfold. And I’m going to tell you exactly what went wrong and why it will happen again unless we fundamentally change how government treats accessible communications.

The question nobody wants to answer

Here’s what the Inquiry doesn’t address: where was the Government Digital Service when emergency communications were being designed? Where were the senior content designers who should have been checking accessibility before materials went live? Why did it take weeks of campaigning by disabled people to get BSL interpretation at press conferences?

That is not necessarily a criticism. The report quotes others who claim to have been sidelined and ignored by Number 10. But GDS exists specifically to set standards for government digital services. They employ some of the best content designers and accessibility specialists in the country. They wrote the service standard. They’ve published extensive guidance on accessible communications. And yet, when the biggest communications challenge in peacetime hit, accessible formats were an afterthought.

So what happened? Two possibilities, both damning.

Option one: GDS failed

If GDS was involved in designing COVID communications from the start and still let inaccessible content go live, that’s a professional failure of staggering proportions. It means years of accessibility advocacy, training and standards were jettisoned the moment things got difficult.

This isn’t how emergencies work in well-prepared organisations. Accessible formats aren’t nice-to-haves you add later. They’re baked into your emergency response templates. Your press conference setup includes BSL interpreters as standard. Your content templates are already designed to work with screen readers. Your crisis communications team includes senior content designers with accessibility expertise.

If GDS was there and this still happened, we need to ask why their voice didn’t carry weight when it mattered most.

Option two: GDS was sidelined

The alternative is worse: GDS wasn’t meaningfully involved at all. When speed became the priority, government departments reverted to their pre-digital ways of working. Press officers wrote the statements. Someone in Number 10 approved them. They went live.

The Inquiry makes clear that decision-making structures were chaotic, particularly in Downing Street. I can confirm that they were from my perspective. In that environment, who’s going to slow things down to check if the content works for people using screen readers? Who’s going to insist on BSL interpretation when there are a dozen other urgent problems?

This is exactly the pattern I’ve seen repeatedly in my career: digital and content expertise matters right up until there’s a crisis, then it’s back to the old ways. The people who know how to make things accessible get pushed to the margins while the urgent takes priority over the essential.

The missing layer: senior content leadership

Here’s what really exposes the problem: there’s no evidence in the Inquiry’s findings of senior content designers having formal oversight of government COVID communications. No sign that accessibility was a gated requirement before materials could go live. No indication that anyone with the authority to stop publication was checking whether vulnerable people could actually understand what was being communicated.

You had SAGE providing scientific advice. You had behavioural science advisers. You had communications teams focused on messaging. But where were the content design leaders ensuring that the brilliant work of translating complex scientific advice into public communications was also accessible?

In a properly structured response, you’d have a senior content designer as part of the core crisis team, with authority to hold back publication until accessibility requirements were met. They’d have checklists. They’d have relationships with disability organisations. They’d have the ear of decision-makers.

Instead, accessibility improvements came later, after disabled people and their organisations fought for them. The Inquiry notes that “improvements made later in the pandemic serve to highlight the difference that proper and timely consideration of accessibility issues can make”. Translation: we proved it was possible, which makes the initial failure even more inexcusable.

Why this matters beyond COVID

It is easy to forget that, before the COVID-19 crisis, the whole of government was focused on preparations for a Hard Brexit. My team went from nine to fifty-six people. Three weeks later, priorities and budgets changed. All those people who had been seconded, all the contractors who had been pulled in from other projects, became redundant overnight. My team went from fifty-six to twelve people. And that set-up never changed throughout the COVID-19 crisis. It seems indicative of how the weight the government gave to the crisis compared to Brexit. 

This isn’t just about one crisis. It exposes a structural problem with how the government treats accessibility.

When I worked on Universal Credit and other government services, I saw this pattern repeatedly. Accessibility is treated as a technical requirement to be checked off, not as a fundamental question about who can access essential services. Content designers advocate for clear language and proper consideration of vulnerable users, but we’re often swimming against institutional assumptions that complexity equals seriousness and that accessibility can be retrofitted. “They’ll work out what the jargon means if they want their benefits,” one senior civil servant told me, when I championed plain English. “They always do.” 

The COVID response laid bare these assumptions. When you’re telling people they can’t leave their homes, that they can’t see their dying relatives, that their livelihoods are ending, you cannot afford for anyone to miss or misunderstand those messages. That’s not just bad practice. It’s structural violence dressed up as emergency pragmatism.

The Inquiry found that “in focusing on how to get messages across to the whole population, the needs of vulnerable groups were sometimes lost”. This is the polite way of saying that vulnerable people were not prioritised. Their ability to understand potentially life-saving information was treated as a secondary consideration.

The AI question: outsourcing accessibility to algorithms

Which brings me to the uncomfortable question: would a future government outsource this to AI?

Right now, there’s enormous pressure to use generative AI to speed up content production. The promise is compelling: faster turnaround, consistency at scale, 24/7 availability. In a crisis, that’s attractive to people making decisions under pressure.

But here’s what AI cannot do:

It cannot understand the lived experience of people who rely on easy-to-read formats because their cognitive disability means complex language creates genuine distress.

It cannot process the living experience of people anxious to claim emergency grants to keep their business afloat in the middle of a global pandemic. 

It cannot know that the absence of BSL interpretation sends a message to deaf people that they don’t matter.

It cannot grasp that someone with aphasia needs information structured in specific ways that current guidelines barely address.

It cannot make the judgement calls about what level of cognitive load is acceptable when you’re telling people their mother is dying alone.

AI could potentially help with some accessibility tasks, for example, checking reading age, flagging jargon and ensuring consistent terminology. But the strategic oversight of accessible communications requires human expertise and judgement. It needs people who understand that accessibility is a civil rights issue, not a technical checklist.

The risk is that AI becomes another way to avoid properly resourcing content design and accessibility expertise. “The AI checked it” replaces “a senior content designer signed it off” and we’re back to the same failures, just with a more modern excuse.

What should have happened

Let me be practical because that’s what’s useful.

In any future crisis, here’s what accessible communications require:

• Emergency response templates that already include accessible formats. Your press conference setup includes BSL interpretation by default, not as something added later when someone complains.
• Senior content designers as part of the core crisis response team, with authority to hold back publication and relationships with the organisations that represent people who need accessible formats.
• Pre-established relationships with disability organisations, so you’re not scrambling to work out what “accessible” means while also managing the crisis.
• Recognition that speed and accessibility aren’t opposing goals. If your emergency templates are already accessible, you can move faster, not slower.
• Training for crisis communications staff that treats accessibility as essential infrastructure, not an optional extra.

The broader failure

The COVID Inquiry’s findings on communications sit within a larger pattern of government failure. They repeatedly found that the government was “too little, too late” in its response. That vulnerable groups were not prioritised. That planning was inadequate.

The failure to provide accessible communications from day one is of a piece with these other failures. It reflects the same assumption that we can deal with vulnerable people’s needs later, once we’ve sorted out the urgent stuff for everyone else.

This assumption is wrong on every level. Vulnerable people are “everyone else”. Their needs aren’t edge cases to be handled separately. And in a genuine emergency, they’re often the people who need clear, accessible information most urgently.

The Inquiry makes clear recommendations about improving accessibility. But recommendations only work if there’s institutional will to implement them and the right people in the room when decisions get made.

The question for GDS and government

So here’s what government digital services need to answer:

• If you were involved in COVID communications from the start, why did accessible formats not go live immediately? What failed in your processes?
• If you weren’t involved, why not? What does it tell us about how the government treats digital expertise in a crisis?
• And most urgently: what’s changed? Because the Inquiry’s recommendations only matter if the next crisis doesn’t repeat the same failures.
• I’m not optimistic. The pattern I’ve seen throughout my career in government is that we learn lessons, write reports, make recommendations, and then when the next crisis hits, we do the same thing again because the structural problems remain unaddressed.

Accessible communications aren’t expensive. They don’t require new technology. They require two things: prioritisation and expertise. Someone senior enough to insist that accessibility happens from day one and someone skilled enough to make it happen well.

The COVID Inquiry has documented the failure. The question is whether anyone with the power to change things is actually listening.

What this means for content designers

If you’re a content designer or accessibility specialist working in or for government, this should worry you. When the biggest crisis in living memory hit, our profession wasn’t in the room where it mattered.

We need to ask ourselves: how do we ensure content design expertise is properly embedded in crisis response structures? How do we make sure accessibility isn’t dropped when things get urgent?

And we need to be honest about the limits of what we can achieve without genuine institutional commitment. All the content design skills in the world don’t help if decision-makers don’t see accessibility as essential.

The Inquiry’s findings are a gift: clear documentation that accessible communication matters and that its absence caused harm. We need to use that evidence to push for the structural changes that prevent it from happening again.

Because it will happen again. The only question is whether we’ll be ready.