Why Kainos is hiring a Data Ethicist

Kainos is hiring its first Data Ethicist to improve trust in AI/ML systems. But why? To do what? And why now?

Photo by Alex Shute on Unsplash

What is Data Ethics anyway?

Data Ethicist. AI Ethicist. Responsible AI Lead. Digital Ethicist. Machine Learning Trust Officer. Machine Learning Trust Programme Manager. Trust and Safety Lead.

On first glance this is an odd collection of job titles. They appear to be philosophical policy-setting roles. But each of these is a real job, with well-known internet-era IT companies Workday, Accenture, Google and TikTok. They are all closely related in their purpose, to ensure the responsible use of AI/ML.

There isn’t yet a “standard” job title across the industry for this responsible adoption of AI, which is evident by the variety of the job titles. Some prefer to talk of “Ethical AI”, others “Responsible AI” and “Trustworthy AI”. In a sense the label doesn’t matter so much because the underlying concern is to ensure that AI systems are fair (not biased), understandable (not opaque), safe (not harmful), secure (not leaky) and have clear accountability (with a person).

At Kainos we have adopted “Data Ethicist” because it best encapsulates the breadth of the role across data and AI systems. Data Ethicists will collaborate closely with Data Scientists, Data Engineers and other data job families.

There has been a rise in the past year or two for dedicated roles to address these unique non-functional concerns because of the increasing reliance on assistance from AI systems and the corresponding rise of high-profile issues. For example,

• Racial bias is accused of facial recognition AI models that haven’t been trained with the right balance of skin colours and racial diversity. Privacy concerns and use for law enforcement are also a challenge for facial recognition systems. If you have time, read the the Alan Turing Institute’s detailed examination of the systemic challenges with facial recognition.
• Regulated industries such as Financial Services have been slower to adopt some AI systems, e.g. deep learning models because of a difficulty to explain AI recommendations. Public issues like the gender bias reported by AI systems for the “Apple Card” demonstrate this challenge.
• Physical safety risks are a priority with Tesla’s Autopilot mode as it develops into Full Self-Driving feature. Media coverage has focused on crashes involving Teslas but it’s clear that one of the most complex computer vision systems will require an equally considered approach to trusted and safe AI.

What skills does a Data Ethicist need?

Kainos is hiring its first Data Ethicist because it wants to ensure its technology is trusted by users. And as we deliver more AI systems it is time to hire the right person to improve our processes, help our customers to succeed and use their strong voice to educate others.

The Data Ethicist will have an interest in following and applying the latest data ethics standards, regional legislation, and practices. But more than this, the candidate will be passionate about going beyond ethical AI research to work within multi-disciplinary software delivery teams to ensure customer solutions are trusted by users. This trust will result from user engagement, improved software development processes, better tooling and tighter QA controls. It will allow even our most ambitious customers to build trust with their users.

Linkedin tells us there are not many data ethicists (in 2022). But this is difficult to quantify given the variety of job titles and lack of standardisation. It is a young discipline that I expect to grow significantly, merging with other forms of data governance, such as data privacy.

Given this is a young discipline we open to hiring someone already in similar role or someone pivoting from a different role. A technology background is an advantage but this can be learned by someone who is curious about technology. We’re interested in someone who can apply their experience to a new discipline, paving the way for others to follow. We are looking for someone experienced in,

1. Developing new processes and practices in an emerging area.
2. Developing risk-based governance frameworks or policy, e.g. data privacy.
3. Simplifying and clearly communicating risks in conversation, documentation and presentation to C-level stakeholders.
4. Critical-thinking and confident engagement with technologists.

This is important now in 2022 for Kainos because we are building more AI/ML with our customers than ever before and we want to ensure our tech benefits peoples lives. We are working in new industries like Defence that has huge ambitions for Data & AI while operating with higher-risk scenarios. We are pioneering new research on Trusted AI with our academic and media partners. And we expect regulation for AI to significantly increase in the next 12 months with the EU AI Act and its conformity assessment based on risk. This is why Kainos is investing now in Data Ethics.

If you are interested in joining the Kainos team as a Data Ethicist you can apply here. Or if you know someone who is a good fit for the role please refer us.