
Maja Kurek
Founder & Principal, RuleBridge Advisory · AIGP · CIPM · CIPP/E
Maja Kurek has spent her career on the receiving end of regulation: close to a decade inside supervised technology organisations that had to make it work. At Google, Photomath and Uber, regulation was never a document but an operating condition.
Her mandates sat where new and evolving rules met live, cross-border operations: converting obligations into processes, controls, documented decisions and evidence that hold up under scrutiny, while the platforms kept running. The result is a particular kind of judgement: which function can actually carry an obligation, what will change in daily operations, what evidence will exist, and where an arrangement will break under pressure.
At RuleBridge, she now applies that operator’s discipline to AI. She advises organisations that build, buy or deploy AI where formal responsibility, real authority and day-to-day ownership most often come apart: in the human roles that review, oversee and answer for AI-supported decisions; when an AI system changes after it has been approved; and in third-party and agentic AI, where authority is delegated and control has to survive a change of provider.
Maja is a member of the European Commission AI Office’s Task Force 2, which is shaping how deepfakes and AI-generated text are labelled under the EU AI Act. She is also a member of its working group developing sectoral good practices. At IEEE, she is a member of the working group developing the P7999 global standards on AI ethics oversight.
Standards and implementation work.
Member, Task Force on Labelling of AI-Generated Content (TF2) · European Commission, AI Office · Code of Practice on Transparency of AI-Generated Content
The task force is shaping implementation guidance for the Code’s labelling measures, including the update of the EU icon, the EU label for AI-generated content.
Member, Working Group on Sectoral Practices · Task Force 2 · European Commission, AI Office
The working group develops sectoral good practices for disclosing deepfakes and AI-generated or manipulated text published on matters of public interest.
Member, IEEE AI Ethics Oversight Working Group (P7999 Series) · IEEE Standards Association
The working group develops global standards for organisational ethics oversight in projects and processes involving AI, for qualifying the individuals who perform it and for certifying organisations.
Selected experience behind the practice.
1. Governance of the human layer of an AI product · Photomath
Behind Photomath’s AI-powered learning product, downloaded more than 300 million times, stood more than 10,000 external mathematics experts: the human layer of the product. Maja scaled that global network from the marketplace’s set-up phase and led relations with it from qualification to offboarding: how changing quality requirements reached ten thousand people in a form they could act on; how their work was priced, rewarded and paid across borders and the data around it protected; what each expert could access, and on what basis; and how disputes were resolved when content assessments, payouts or suspensions affected people’s access to work and income.
2. Accountability through acquisition and provider change · Photomath → Google
Following Google’s acquisition of Photomath, an operation that already worked had to be carried into the control environment of one of the most heavily supervised technology companies in the world, without being interrupted. Maja led her team’s integration, coordinating changes to approval paths and compliance processes. She also led the operational implementation of EU consumer-protection requirements, including changes under the New Deal for Consumers, for an AI-powered app used mainly by school students, parents and teachers: transparency, contractual rights and access to remedy. When the Photomath platform and its expert network moved to a new provider, she was responsible for the part of the transition that affected the experts: communication with the global network, data and access requirements, the handover of responsibilities to the vendor and outstanding expert payments.
3. Regulated platform adaptation · Uber
At Uber, Maja worked inside a business as its regulatory environment moved from contested status to formal regulation. As transport, tax and labour requirements arrived and changed, she implemented the compliance processes through which an ecosystem of major independent fleet partners was brought into line while the rules were still moving, and led the implementation of a cross-functional programme for the tax and labour changes, in plans that major fleets and micro-companies alike could carry out. The work meant translating each obligation into partner-facing processes, payment compliance, vendor service levels and monitoring routines that had to hold at scale, under public and regulatory attention, without interrupting the service.
Across these environments, the recurring challenge was never simply to interpret a requirement, but to make responsibility workable across functions, systems and external parties: controls over data and access, payment safeguards, vendor and third-party accountability, documentation, evidence and continuity, and to keep it working through change. Frameworks name these controls; she has operated them. That is the operating layer from which Maja approaches AI governance.
Selected publications.
The Vendor’s AI Model Did Not Change. But Its Risk Did. · Article, Corporate Compliance Insights
In Corporate Compliance Insights, she examines how changes in the human roles behind an AI service can change third-party risk while the model stays the same. The article develops practical implications for the approval and continuing oversight of external AI services.
AI Change Accountability & Reauthorisation · Method
Practitioner-derived, organisation-side method that Maja developed to help organisations retain decision authority as AI systems and their conditions of use change. The published method brings material changes back to an accountable decision on continued use, with responsibility for implementation and evidence of completion.
Governing AI After Deployment: Field Evidence on Model Changes, Platform Configuration and Organisational Authority · Working paper, SSRN
Her field study of a real AI consolidation initiative, conducted with the organisation’s actual decision-makers. It contributes field evidence on how post-deployment AI change travels across provider, platform, configuration and organisational authority boundaries, and what must connect technical visibility or rollback to a current, owned and evidenced response.
Human Oversight Reliance Review
She also developed the Human Oversight Reliance Review, a published method for establishing what a human oversight arrangement can actually support, and returns the question to a decision-maker when material conditions change.
- AIGP · Artificial Intelligence Governance Professional (IAPP)
- CIPM · Certified Information Privacy Manager (IAPP)
- CIPP/E · Certified Information Privacy Professional / Europe (IAPP)
- Master’s degree in Business Economics, Management
At RuleBridge, that experience is applied to one discipline: operational accountability for AI.