Every business now runs on AI.
Few are in control of it.
GenWA™ puts AI to work, safely: as advice, as a product, or as an agent organisation we incubate or accelerate for you.
Animation: a chomper chases stray AI tools around the GenWA flywheel and eats every token and a running total of token burn counts up by £1,000 for each token eaten; the screen reads game over, you are out of tokens; then the ring lights up.
of employees have used AI tools their employer never approved.
of employers have an AI policy at all.
of leaders can only partly see what AI costs them.
of businesses using AI have security practices that cover it.
GenWA gives you visibility and evidence across the AI your people use, and control over the AI used through GenWA. Any model. In your AWS or Azure account, or hosted by GenWA on AWS as agreed.
Ninety seconds on AI control.
Music and sound design only. What the film says is on screen, and a text version follows.
Read what the film says
Right now, someone in your business is using AI. Only 31% of employers have an AI policy at all (CIPD, 2026), and 42% of leaders can only partly see what AI costs them (KPMG UK, 2026). Every business now runs on AI. Few are in control of it. Do you know how your organisation is using AI today? Most can’t answer.
Nobody owns the whole picture, and that costs you four ways. The bill you can’t see. Paying premium by default: the premium share of AI cost rose from 6% to 56% in ten months (Ramp, 2026). Data you don’t control: only 24% of businesses using AI have security practices that cover it (DSIT, 2026). And nothing you can prove.
And the rules are catching up. A regulator will ask, and so will your next tender: UK GDPR as updated by the Data (Use and Access) Act, ICO guidance on AI, ISO/IEC 42001, and FCA SYSC and the EU AI Act where they apply.
The answer isn’t to ban it, or to add another tool. Treat AI like any other part of the business. See it. Control it. Prove it. Put it to work. GenWA is a control layer between your business and every AI lab: your rules on every request, whichever lab answers, set up with your team in your own cloud account.
Five products, inside the control harnesses: Analyse, Guard, Evidence, Automate and Grow. Start where it hurts. Add the rest when you’re ready. What good looks like: an estimated 56% of AI spend saved for the same work, in our cost model for a 300-person business, through the right model for every task, unused seats reclaimed and duplicates retired. Your AI health check shows your own figure.
Built by people who ran FCA-regulated businesses, and know that control is what lets you move fast. Find out where your AI stands. Then decide. Start your AI health check: under 60 seconds online, first findings within 48 hours, no obligation.
Start at the right place for your business.
See what good looks like, what to avoid, and three ways to begin, from a training session to a full plan.
Start hereAlready using AISee it, control it, evidence itSee the tools and the spend, enforce your rules on the AI used through GenWA, and keep a record an auditor can read.
See the productsReady to build with AIThe Incubator and the AcceleratorAn agent organisation we run for you in a controlled environment, or senior AI leadership time if you are a founder-led business.
Explore the IncubatorSee how it works in your sector.
Each sector page shows where AI is already at work, where it leaks, and the agents we suggest.
Financial servicesEvery customer message needs a record. Every decision needs a name.
InsuranceThe most AI-intensive part of UK finance, and the most exposed.
ManufacturingAI in the back office, with engineering IP in the prompt.
TechnologyThe highest AI adoption of any sector, with engineers calling models directly.
AccountancyManual work everywhere, AI in strategy, little in control.
Real estateLeases, valuations and tenant data are going into AI tools. Who checked the output?
Five products, one control layer, inside control harnesses.
See it, control it, prove it, put it to work and scale it. Select a product to explore it.
What AI control means for you.
Advice, products, an organisation we run for you, an investment in you, or a build.
Most businesses start with one and add the others. All five sit on the same control layer, so nothing you do with us has to be undone. It is built for businesses with no in-house AI team.
Independent AI audit, policy and controls, training, strategy and planning. Fixed fees, credited against year one if you go on to the products.
See advisory services02ProductsAnalyse · Guard · Evidence · Automate · GrowYour rules, any model. Then agents on top of them.The control layer between your business and every AI lab, and a trained agent team deployed on it. Modular, priced as a range configured to your requirements, run in your cloud or ours. Built for a 10-person business as much as a 5,000-person one.
Explore the products03IncubatorAn agreed engagement · Paid per outcomeWe build and run an agent-first function, team or organisation for you, in a controlled environment. Agreed with you, no application, paid per outcome.A function, a team or a whole operating unit, run agent-first through the GenWA methodology with a small human core, live in twelve weeks. You pay for verified outcomes, with a cap you set.
See the IncubatorSenior AI leadership time, advisory and platform access, paid for by equity, a project fee, a retainer or a licence. By application only, with ten places a year.
Bespoke software for a business-critical mission, led by a senior engineer with an agentic team and control built in. Scoped and priced in writing.
The right model for every job. Stop paying premium for routine work.
We know which model does each job well at the lowest cost, and we set it once. Cut your token burn. Keep your momentum.
How model selection worksThese are averages. Your health check uses your own usage.
See how it is calculated
Spend: people × AI spend per person. The default of £92 a month comes from 15% of software budget going to AI (Ramp, 2026) on £7,400 software spend per employee (Zylo, 2026). Savings are applied in turn, so they never double count. Wrong model: 56% of AI spend is on premium models (Ramp, 2026), three-quarters of it routable to standard models at 85% lower cost (RouteLLM, 2024). The other three are GenWA working assumptions, not published statistics: large-context sessions, 20% of what remains, half of it avoidable; repeated documents and context, 25% of what remains, cut by 75% through caching, which providers bill at a fraction of the standard input price; duplicate tasks, 8% of what remains, 90% reusable. The total saving is capped at 70% of spend. Your health check replaces every assumption with your numbers.
The questions boards ask us.
What is GenWA and what does it do?
GenWA helps UK businesses see, control and evidence the AI they already use, then put AI to work inside rules they set.
We start with an AI health check, then offer five products: Analyse, Guard, Evidence, Automate and Grow.
A named person in your business owns every consequential decision. See how we design AI to work as one organisation.
How can I find out what AI my staff are using?
Start with our AI health check. Answer a short questionnaire, upload a usage export or book an on-site visit. The usage-export check runs locally in your browser with no AI API calls. We send first findings within 48 hours, then the full audit and a 90-day plan within ten working days.
Who is GenWA for?
GenWA is built for UK businesses already using or adopting AI, especially those of 10 to 500 people without an in-house AI team. Owners, CEOs, CFOs, COOs, CTOs and compliance leads use it to see AI use and cost, enforce rules and keep evidence. Start-ups and enterprises have routes too.
How much does an AI audit cost?
Indicative audit fees, excluding VAT, are 7,500 pounds for 50 to 250 people, 12,500 pounds for 250 to 1,000 people and 19,500 pounds for enterprise. Fees are confirmed in writing before you commit, and the audit fee is credited in full against year one of a platform plan.
Where does GenWA run and which AI models can it use?
GenWA runs in your own AWS or Azure account, or in GenWA's AWS as agreed. It works with approved providers: OpenAI, Anthropic, Amazon Bedrock, Azure OpenAI, Google Gemini and private models. Personal data is redacted before approved external model calls. Data processing terms are confirmed in the security review.
Does GenWA work alongside Copilot, ChatGPT and the AI tools we already pay for?
Yes. You keep the tools your people use. GenWA puts your rules, redaction and metering in front of connected, routed integrations, so their use appears in one register and one cost view. Coverage depends on the integrations scoped in your audit, and we do not claim to discover every unmanaged app.
We prepare the evidence; certification is issued by accredited bodies.
What seasoned operators are saying about AI.
Three practical reads from GenWA's leadership team on what AI costs, how to keep it in control and how to tell whether it is paying back.
By Steven Myers · 5 min read
Guard
Harnesses: the controls that make AI safe to put to work
What a harness is, what it does, what happens without one, and why deploying it well is a skill of its own.
By Lukasz Piotrowski · 5 min read
Advisory
Why AI returns stall, and how to set AI up properly
Many AI trials never show a return. A five-step frame and the measures that tell you if AI is paying its way.
By John White · 5 min read
Find out where your AI stands. Then decide.
First findings within 48 hours of the health check · Audit fee credited against year one