On this page
Questions you could ask me
Between approximately 2016 and 2020, a substantial part of my work concerned website security, data protection and compliance.
The introduction of GDPR, which became applicable in May 2018, brought new urgency to some of these issues.
But my interest wasn’t limited to satisfying regulatory requirements.
I wanted to establish what was wrong, address the problems and provide meaningful evidence that the work had made a difference.
Starting with an assessment
A recurring feature of my security engagements was an initial audit.
Before making changes, I wanted to understand the current condition of a website and the issues that needed attention.
The remedial work followed from that assessment.
That might involve correcting identified weaknesses, improving configuration or addressing other technical and procedural problems.
The important point was that the work was directed by the evidence available, rather than by a generic assumption that installing a particular product would make everything secure.
Afterwards, I used third-party assessment applications to evaluate the results.
Customers could inspect those results themselves.
That independence mattered to me.
I could explain what I had changed, but customers shouldn’t have to rely exclusively on my assessment of whether the changes had been effective.
Recovering compromised systems
I also worked on websites that had already been compromised.
That presents a different kind of challenge.
Rather than beginning with a system that appears to be operating normally, you are dealing with something whose integrity has already been called into question.
Recovery requires care. Restoring visible functionality isn’t necessarily enough if the underlying problems remain.
The work reinforced the importance of understanding the condition of a system before making assumptions about its reliability.
It also highlighted the need for a disciplined relationship between assessment, remedial action and verification.
Security is not a single change made once.
It is an ongoing responsibility.
GDPR and the wider picture
GDPR brought additional attention to the ways organisations collected, used, stored and protected personal information.
Technical security was an important part of that picture, but it wasn’t the whole of it.
Organisations also needed to consider their processes, responsibilities and the reasons for handling information.
My work increasingly involved helping customers understand these wider requirements alongside the technical measures necessary to support them.
Over time, the engagements became more substantial.
Before COVID changed my immediate priorities, I was working with a couple of companies towards ISO/IEC 27001 information-security compliance.
That represented a move from individual website assessments towards more comprehensive organisational information-security management.
The work concerned preparation and improvement; I wouldn’t equate that with a claim that those organisations had necessarily achieved certification.
Applying the experience elsewhere
Later, while undertaking other technical responsibilities at a company in Norway, I introduced information-security practices informed by ISO 27001.
That work drew on experience accumulated through years of security assessments, remedial projects and compliance activity.
It also reinforced something I had come to appreciate: security is not merely the responsibility of whoever maintains the servers or applications.
It affects the way an organisation understands risk, assigns responsibilities, manages information and makes decisions.
A sound technical control can be undermined by an ineffective process.
An apparently sensible policy may achieve little if nobody can demonstrate that it is being followed.
Evidence rather than reassurance
Security is an area where absolute statements are particularly dangerous.
A successful assessment doesn’t prove that a system can never be compromised.
A favourable result from a third-party tool doesn’t amount to comprehensive assurance.
And a documented policy doesn’t establish that a control is effective in practice.
But meaningful assessments, recorded improvements and repeatable verification are still valuable.
They allow people to make better-informed decisions.
They also provide a basis for identifying what remains to be addressed.
That is the principle I carried forward from this period of work.
Good security practice should produce more than a feeling of reassurance. It should produce evidence that can be examined, questioned and acted upon.
Project context
Research programme: Making data analysis understandable, reproducible and accountable
Topic: Data protection
Explore further
Further reading is listed below.