AI is getting better at producing answers.
Your organisation still needs people who know when to trust them, and when not to.
I help organisations adopt AI responsibly, understand what AI is doing to human judgement, and then build the conditions for people to use it well.
Cyberpsychologist. Founding architect of the NHS AI Taskforce. AI Ambassador at the Department of Health and Social Care. 25 years in NHS communications and digital transformation.
Confident AI use is not the same as competent AI use.
Your people may be using AI fluently. They may be producing work faster and completing more of it. Your adoption figures will show all of that. They will not show whether people can still recognise when AI is wrong, explain the decisions they have made, or work well without it.
The distance between confident and competent use is the Judgement Gap. It is a property of the person, not the tool, and it shows up later, in a decision nobody double-checked.
That is the problem I work on.
Why it matters
AI can be a powerful tool, but it also introduces new risks.
When judgement is compromised, the consequences are real.
Verification declines
AI becomes the first answer rather than the second opinion
Expertise becomes harder to see
People can produce competent-looking work without being able to explain the reasoning behind it.
Oversight becomes performative
Someone remains in the loop but no longer has the time, conditions or confidence to challenge the system.
Governance gets ahead of behaviour
Policies say what people should do. They do not tell you what happens in practice.
A QUICK REFLECTION
Where is judgement actually at risk in your organisation?
Four questions to ask before you spend anything on training
Could you explain how that AI assisted decision was reached?
1.
Would it survive scrutiny?
2.
Did you verify the output, or trust it because it’s usually right and it sounded OK?
3.
If your most experienced person left tomorrow, how much of their judgement now sits inside a tool no one is checking?
4.
If you are not sure of the answers, that uncertainty is the finding. It is worth a conversation.
How I can help
ORGANISATIONS
Diagnostics, advisory and programmes designed to support responsible, judgement-led AI adoption.
EXECUTIVES
Strategic support for leaders who need to understand the behavioural consequences of AI adoption.
PROFESSIONALS
Practical learning for people who want to use AI competently without losing their judgement.
SPEAKING
Talks and workshops on AI, trust and professional judgement, for conferences and internal events.
I work with NHS and public sector bodies, professional services and any organisation where AI-assisted decisions carry consequences.
“Sonya’s framework was the missing implementation piece in our governance strategy.”
Chief Information Officer, GoJ
START WHERE IT’S USEFUL
A first conversation, not a sales call.
If you are bringing AI into your organisation and want a clear-eyed view of what it costs in human terms, this is the right place to start.