Chatbots, copilots and AI agents fail in ways a model validation will not catch: confident wrong answers, leaked data, manipulated prompts, silent vendor changes. Pick one GenAI system, answer twelve questions, and see whether it is ready to launch, what must be fixed first, and how your controls compare with the practices the FCA highlighted in its October 2026 AI review.
Choose the closest match. It changes which questions are treated as critical, so a gap that would be tolerable in an internal tool can block the launch of a customer-facing one.
Two questions per area. Answer for this one system as it would run on launch day. Questions marked Critical are launch blockers for the system you picked.
A PDF of your results to share with your team: your score by area, every gap with what good looks like, the FCA review points, and the regulatory touchpoints for this system. Leave your details and it downloads straight away.
LINXS Advisory works with banks and financial firms on GenAI launch readiness, evaluation and red-teaming plans, use-case governance, and the line between model risk and AI risk. To sort your whole AI inventory and assign each system the right controls, see our AI Inventory Sort.
Get in Touch Try: SS1/23 AI Model CheckThe check runs in your browser and nothing you answer is stored or sent unless you request the report. Results are indicative, draw on the FCA’s October 2026 review AI in financial services: good practice and areas for improvement, which creates no new requirements, and are not regulatory or legal advice.