Six checks. A clearer standard.
What does it mean for a business to be good to go with AI? Our answer is six practical checks, supported by evidence from the way people actually work.
01 / Visible use
We know how our people use AI.
Have in place: A current register of tools, use cases, teams, data and significant decisions affected by AI.
Check it works: Sample staff answers match the register; previously unknown tools have a route into it.
02 / Clear ownership
Someone is accountable and knows what to do.
Have in place: A named accountable leader, operational owner, deputy and escalation route.
Check it works: The owner can explain how a new tool or concern is handled, including when they are absent.
03 / Practical boundaries
Our people know what is allowed.
Have in place: Approved tools, task-specific data boundaries, acceptable-use rules and an approval process.
Check it works: Staff can distinguish acceptable use from cases that need permission or must stop.
04 / Human control
People check and own important decisions.
Have in place: Verification and sign-off requirements matched to the consequences of getting an output wrong.
Check it works: A real example shows a person checking evidence and taking responsibility for the result.
05 / Prepared people
Our team can use AI safely and usefully.
Have in place: Practical training, accessible staff guidance and onboarding for new starters.
Check it works: Staff can spot a plausible error, protect confidential information and explain when to ask for help.
06 / Ongoing review
We keep it working as AI changes.
Have in place: Review dates, incident records, change triggers and a maintained evidence trail.
Check it works: The next review is scheduled and a new model, tool or incident can trigger an earlier check.
What counts as meeting the standard?
Define the teams, tools and uses being assessed. All six checks need appropriate evidence for that scope, with critical risks resolved or the affected use stopped. Record the assessment date, any scope limitations and the next review. Readiness can change when a tool, workflow or risk changes.
GOOD TO GO is our consultancy’s practical readiness standard. It is not government certification, ISO certification or a guarantee of legal compliance or freedom from incidents.
A practical foundation, informed by recognised guidance
The UK Government’s published AI Management Essentials consultation tool covers organisational processes such as accountability, risk management and ongoing monitoring. Our checklist translates related management principles into simple questions for staff AI use; it is not an official AIME assessment. Read the published AIME tool.
Set a date—and work backwards
We recommend that UK SMEs aim to have these foundations in place by summer 2027, sooner where exposure calls for it. This is our recommended planning target, not a statutory deadline. Start now by naming an owner and recording current use; do not wait for a date to address a known serious risk.
Need help establishing the evidence?
The Review identifies your gaps and next steps. Governance helps implement and test the foundations. You can also download the free four-page briefing.