AI privacy engineer Modupe Akintan argues that protecting people’s data needs to be part of how AI systems are built and run, not a promise added afterwards. Her work offers a concrete view of a growing challenge: turning legal and policy requirements into technical controls teams actually use.
Watch Desk analysis
What happened
In a Digital Journal profile published on 15 October, Akintan describes privacy engineering for AI as the design of technical controls, risk assessments and standards to protect personal data. She says the aim is to catch risks before systems move from prototype to production, and to turn legal requirements into constraints in code, contracts and everyday workflows.
Akintan works as a Privacy and AI Engineer at an unnamed technology company and contributes to the Cloud Security Alliance’s Data Privacy Engineering Working Group. The article also describes her work on AI governance and privacy-by-design guidance. Its central case is her account of the field, rather than evidence that a particular company has changed its practices.
The profile includes a pointed counterargument from an unnamed policy analyst, who warns that “digital confidence” can become compliance theatre if organisations publish frameworks while leaving data practices untouched. Akintan agrees that titles and working groups are no substitute for hard choices.
Why it matters
AI systems can process personal data at scale, but a policy document alone cannot constrain what a product does. Akintan’s argument is that privacy requirements should shape the system’s design and daily operation, with clear rules and accountability when things go wrong. That puts the emphasis on engineering choices readers can ask about, not reassuring language.
The article gives a practitioner’s perspective, not an independent assessment of how widely the field is adopted or whether its methods work in practice. Still, the contrast it raises is useful: a governance framework matters only if it changes what teams build and what data systems are allowed to use.
Our read
The best line in this debate is also the least comfortable for a marketing team: if users have to offer blind trust, the work is not finished. Privacy engineering is a practical idea when it produces enforceable controls, not just another badge on the product page. Ask what changed in the code, the data flows and the accountability process.
What to watch
- Whether Akintan or the Cloud Security Alliance publishes practical AI privacy controls teams can adopt.
- Whether organisations show how privacy requirements alter product design and data handling.
- Whether regulators or independent researchers assess how these practices work in deployed systems.
Discussion spark: Should AI companies have to show that privacy rules are enforced in their systems, rather than simply publish governance frameworks?
Sources and evidence
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.