This article was written and edited by Joshua (Isidore) Foakes (LinkedIn) and Jen Stirrup (LinkedIn, Get in touch with Jen directly here).
With the Pentagon’s Joint Artificial Intelligence Center (JAIC) recently creating a new position for AI Assurance, and many similar roles being offered on employment websites, a new trend is emerging for the hiring of Chief Ethics Officer roles for AI in various organisations. The future of this position is impossible to gauge but it is already having an impact on the leading AI hubs of the world.
A Chief Ethics Officer works with all levels of an organisation to develop and distribute ethical codes and training programs for employees, as well as monitor and audit compliance within the company. This role has expanded and become more specific over the years, as the need for showing ethical compliance within companies has become more and more apparent for companies to succeed. Government legislation such as the Data Protection Act (2018) in the UK and the UK General Data Protection Regulation (UK-GDPR) has been instrumental in this regard in helping to focus minds on the use of data and its impact on individuals. As organisations become more educated in their responsibilities towards individuals, this has begun to solidify the meaning and necessity of the role in many organisations.
Why are Chief Ethics Officers Needed?
Yet the question remains exactly why a Chief Ethics Officer is needed in businesses. Pressures to stay competitive lead to a resistance by some to adopt new ‘bureaucratic’ roles which may slow down revenues in the short term. In contrast to fields such as medicine and law, which are closely licensed by governments with regulated ethical principles, machine learning has no safeguard despite having great potential to harm others and work against the common good.
Nonetheless, ethical considerations are part of any company’s social responsibility. Beyond the direct motive for profit, ethical self-regulation can show itself a proof of good faith towards the community and help with its social commitments. With this in mind, a chief ethics officer is valuable, not only to ensure that all parts of the company keep this social responsibility in mind, but also to make sure that approaches are consistent with one another, and reflective of the same goals.
Chief Ethics Officers are being increasingly found across the financial and health sectors, and the tech industry is making strides towards its own solution to the various problems unearthed over recent years. Facebook has set up an ethical review policy, Google consults philosophers on various questions such as Artificial intelligence, and Microsoft has its own ethical researchers. The trend is clear, with more and more dedicated roles cropping up on the question of ethical company policy.
The AI Chief Ethics Officer
But why a Chief Ethics Officer for AI? The development and use of AI has many ethical considerations which continue to be worked out today. Principles and action areas continue to be developed along the lines of transparency, inclusion, responsibility, impartiality, reliability, security and privacy. Just as an Ethics Officer revolves around bringing different departments together to consult on following ethical guidelines in practice, the AI Ethics Officer is a specialised role that requires helping technology departments and business leaders deal with the correct way to apply AI.
Algorithms must be kept free from bias, not infringing upon the rights of others for unfair gain or reflecting explicit or implicit biases in the programmer. Since we are often blind to our own biases, a Chief Ethics Officer remains a dedicated neutral party to correct these mistakes and ensure the use of AI in the company is impartial, reliable, and secure. These areas of AI implicitly involve all areas of the development and deployment process and as a result likely involve the whole IT department.
This touches on the reason we need Chief Ethics Officers in the first place. Some criticisms of AI Ethics Officers may argue that it is the duty of the entire organisation to ensure ethical AI standards are met. This is certainly true, as acting morally is the duty of everyone. However, beyond even the advantage of CEOs managing ethical standards and providing a consistency in how they are applied across the company, AI Ethics Officers will be needed to understand and communicate the purpose and mechanisms of the algorithms used themselves, and require a more specialised role to ensure the same consistency is kept.
Notwithstanding, the situation without considering fair use of AI seems bleak. Hollywood hypes up the potential of AI sentience emerging and leading to the subjection of the human race, but far less theatrical, and perhaps even more worrying, are the issues around AI being deployed in the wrong hands to kill people without human input, to destroy livelihoods by misidentifying suspects or dehumanising them by failing to recognise people of colour as human.
Data breaches often occur in companies that fail to secure their data or misuse them, even when the algorithms used are not perpetuating the biases of the programming team: more than 20 billion data breaches occurred in 2020 alone, all of which have an effect on lives and livelihoods of those whose privacy has been compromised. It is not the case that a few small organizations have acted in a roguish manner; even in 2022, the UK Ministry of Justice has fallen foul of the ICO in failing to provide, without undue delay, 7,753 data subjects with a copy of their data in accordance with its legal obligations.
Challenges of the Role
This is not to say that ethical considerations stop simply at a new hire, or that the challenges involved with installing a new role in the company are trivial. According to an article by David D. Dick on the subject, among these challenges, friction between the new Ethics Officer and other heads of departments is almost a given: when recommendations may be costly to implement, Ethics Officers may be opposed by the CFO, and where it may lead to legal risks (such as when recommending statements of apology and admission of fault) these recommendations may be opposed by general counsel. Moreover, it remains to be seen how in demand a new company role may be for the position, since university roles with their greater independence and freedom would be more attractive to qualified ethicists.
None of these challenges are insurmountable, however. Part of the role of a Chief Ethics Officer is managing these competing voices in the company, in order to produce a policy consistent with the need to avoid ethical pitfalls and to make each branch’s approach complementary from a single policy foundation. Whereas it would be rash to promise a perfectly harmonious ethical policy with the motives of other branches of the company, stable improvements can be made through the direction of an ethics officer over time, yielding continuous growth to the company.
The challenges for an AICEO go further: technical AI knowledge and perception, ethical considerations, social science and technology law familiarity, and business strategy know-how are all necessary parts of an AI Ethics Officer’s qualifications, since this multidisciplinary knowledge is needed for the interconnected nature of AI. With adequate familiarity in these disciplines, effective and inclusive governance, and a company-wide approach to achieving these goals, the appropriate education and training of employees can be managed to achieve the proper goal of the Chief Ethics Officer: to establish the proper ethical compliance of all parts of the company to fulfil its social responsibility and forge a stable path of growth for it.
What’s next?
Organizations make increasing demands on customers to give us their data, but are still falling foul of the law. When a customer gives you their data, there is an expectation that the company will hold the data sacred, and will offer better service since they understand the customer better. Unfortunately, they are often failing to deliver to customer expectations in a way that is suggested when people hand their data over to companies. The Chief Ethics Officer certainly has a difficult job to do, and it is exciting to see where this role will go in the future.