Good Morning Reader it's Maria,

I came across a post this week that stopped me mid-scroll. Sol Rashidi, the world's first Chief AI Officer back in 2016 (and a woman), was speaking at the AI Congress about something that should concern all: We are deploying one of the most powerful technologies in human history, without any security around it, not even the equivalent of seatbelts.

I think about this a lot.

I am genuinely for AI. I have built my work around it, and I do believe it is going to empower us to do extraordinary things. But that doesn't mean that I'm wearing pink glasses and not seeing the risks of what could happen when something this powerful is deployed without proper guardrails.

And right now, that is exactly what is happening.

If podcasts are more your speed, I've got you covered; there's a discussion about this topic available now here.

THE BIG IDEA

Accountability is not the enemy of innovation; it is what makes it last.

Here is the thing. I have spent the last few years building my entire work around AI. I believe in it. I have seen what it can do when people use it well. So when I say what I am about to say, I am not coming at this as a sceptic. I am coming at it as someone who thinks we deserve better than what we are currently doing.

We would not let a new drug reach the market without clinical trials. We would not allow food to be sold without knowing what is in it. We would not let a car be driven on public roads without basic safety standards. These rules did not kill those industries. They gave people enough confidence in them to let them grow.

And here is my favourite example of what happens when the world actually gets this right.

In the 1980s, scientists discovered a massive hole forming in the ozone layer over Antarctica, caused by chemicals called CFCs, found in everything from refrigerators to hairspray. Nations came together, agreed on the Montreal Protocol in 1987, and held industries to it. It required real sacrifice from sectors that had been profitable for decades. And it worked. Today, over 99% of those substances have been phased out and the ozone layer is healing, on track to recover to 1980 levels by the middle of this century.

The world saw a serious problem, agreed on rules, and followed through.

We can do the same with AI. The question is whether we will move fast enough.

Europe is trying. The EU AI Act, which reaches full enforcement in August 2026, is the most serious attempt so far to bring real structure to how AI is built and used. It categorises systems by risk level, sets clear obligations for organisations deploying them, and includes actual consequences for non-compliance.

A lot of businesses heard that and filed it under "something for the legal team." I get it. But that is the wrong call.

Because what the Act is really asking is a question every leader should already be sitting with: do you actually know what your AI tools are doing, and who is responsible when something goes wrong?

Most organisations cannot answer that because it has never been part of how we think about rolling out technology. You find something that works, you use it, you move on. The idea of documenting what data a tool uses, understanding its known limitations, or naming someone accountable for its outputs feels like overhead.

It is not overhead. It is basic professional responsibility.

Everybody is talking about AI accountability right now. Not many are actually doing something about it. The EU is, and I hope the rest of the world follows.

And in the meantime, Reader, there is plenty you can do in your own business without waiting for anyone to make it law.

Don't know where to start?

​Book a free consultation and let's chat!

THE ACTION STEP

Have an AI accountability conversation with your team this week.

It does not need to be a formal meeting 30 minutes over coffee will do. The point is to surface what is currently invisible.

Put these four questions on the table:

1. What AI tools are we currently using, officially or unofficially?
2. What data are we feeding into those tools, and do we know where that data goes?
3. If one of these tools gave us a wrong or biased output, would we catch it? Who would be responsible?
4. Does anyone on our team own AI accountability, or is it currently nobody's job?

One more thing before you gather everyone: declare an amnesty. Whatever anyone has been using on the side, however unofficial, this is the moment to put it on the table. No eye-rolls, no telling-offs. You want the full picture, and you will only get it if people feel safe enough to share it.

You do not need to have the answers yet. The conversation is the starting point. Most teams have never had it.

Have you just signed up? See all previous newsletters here.​

AI MADE SIMPLE

Use AI to do a quick risk scan of your current tools.

Once you have your list from the conversation above, take it to Claude or ChatGPT and try this prompt:

"Here is a list of AI tools my team currently uses: [paste your list]. We are a [describe your business and industry]. Can you flag any potential risks I should be aware of, including data privacy concerns, bias risks, over-reliance issues, or anything relevant to EU AI Act compliance? Keep it practical and specific to our context."

You will get a useful starting point in seconds. It will not be a legal audit, and treat it as a first pass rather than a final answer. But it will show you where to look. And it will almost certainly surface something your team has not discussed yet.

That is the point: Use AI to help you use AI more responsibly.

That's all for today Reader

Have a great weekend!👋🏼

Take Care

Maria

PS: If you want to explore what working together looks like, book a free call

PPS: If you enjoy these emails and want to do something nice, you can buy me a coffee 😉

​​​