At first glance, digital ethics boards look like the perfect response to the technology sector’s ethical imperative with the recent spate of companies setting up such initiatives. These boards help businesses to set policies governing their use of technology and to draw out ethical issues they may have overlooked.
Businesses adopting variations on the theme include Microsoft, IBM and Twitter, as well as many smaller tech-sector companies. But their work on digital ethics boards is not offering the universal immunity to accusations of unethical attitudes that they had hoped for, and is proving controversial in its own right.
How do digital ethics boards work?
As a concept, a digital or AI ethics board is a straightforward proposition. It operates as an advisory board or committee to the business, usually with a majority of external advisers, which the business appoints for their expertise in particular areas of AI or digital. There may also be representatives of the business itself.
The board operates as a sounding board, bringing an outside perspective on business plans and AI and big-data initiatives. Leadership may consult it on anything from cultural attitudes to big-data use to helping product and service design to identify potential AI pitfalls.
That all sounds uncontroversial. The business gets an expert team of advisers to recognize and negotiate key ethical challenges, thereby protecting a wider group of stakeholders from potential harm. As an additional positive, the business is seen to be taking ethical issues seriously – it can point to its digital ethics board as evidence that it is engaged with any external challenge and of its eagerness to do the right thing.
But there are real difficulties here, too. One question is whether board members are genuinely independent. In most cases, they are paid for their time and expertise – in which case, will they be willing to bite the hand that feeds them?
In addition, there is the issue of transparency. Should the board’s deliberations be shared publicly, risking embarrassment for the business, or strictly controlled using non-disclosure agreements, for instance. The latter course leaves the business vulnerable to accusations of secrecy, meaning that observers may assume unethical activity is happening behind closed doors, with the ethics board simply a token to appease critics.
Questions may also arise around the degree of influence a digital ethics board wields over the business it advises. No business can give a right of veto to an advisory board largely composed of external figures, but, inevitably, views will differ from time to time between board members and company leadership. Companies have the delicate task of making board members feel their opinions are valued and considered (keeping them onside and projecting a positive image to the public). while not allowing them to de facto control the business.
Learning from the winners – and losers
Technology and weaponry manufacturer Axon launched its AI ethics board in 2018, with a brief to advise the business on the ethical issues around the use of AI technologies in policing. Initially, the board worked closely with the company, prompting Axon to rethink its approach to several new ventures. However, earlier this year, Axon announced that, as part of a strategy to tackle mass shootings in schools, it was developing drones equipped with Tasers and real-time surveillance tools. Nine members of its AI ethics board subsequently resigned, claiming they had urged the company not to proceed with the proposal and accusing it of seeking to exploit commercially the tragedy of school shootings in Uvalde and Buffalo.
Some companies have run into problems right from the start. In 2019, Google announced it was shutting down its digital ethics board, which it had set up to consider issues around AI, machine learning and facial recognition, within weeks of its launch. The company had been widely criticized for appointing a conservative think-tank executive to the board, with critics claiming she had shown prejudice against LGBT and immigrant communities.
Making a recipe for success
This is not to suggest that there has been no successful deployments. Pharmaceuticals giant Merck has been widely praised for its decision to set up a digital ethics advisory panel to guide it through issues around data usage and digital innovation. The initiative reflects the pharmaceuticals sector’s increasing use of AI and big data in drug discovery and medical research.
What, then, are the key ingredients of a successfully deployed digital ethics board? The answer, broadly speaking, lies in anticipating potential difficulties in advance and putting in place effective conflict-resolution structures.
This starts with imbuing the digital ethics board with clarity of purpose, so there is no confusion about its remit. Board’s members should be absolutely certain about where their responsibilities begin and end.
Equally, digital ethics boards require agency to do their work effectively. They need access to sensitive information and documentation, and to the people within the business who are relevant to the issues they are examining. There must also be channels to facilitate engagement with the business’s senior leadership.
Businesses can also prepare for confrontation so that it can be effectively defused. Boards cannot expect that their advice will always be followed to the letter, but there should be potential for bilateral discussion with leadership when areas of disagreement arise. At Merck, for example, in the event that the company’s leadership chooses not to follow the board’s advice, governance rules require the company to provide a written statement explaining its decision.
Prepare to succeed
In other words, setting up a digital ethics board is not something that can be done without considerable thought and preparation. However, companies often fail to put in that work in advance – often because they’ve launched an ethics initiative in response to a particular controversy, rather than as a proactive, structured, and forward-thinking measure.
An alternative – and one that avoids all these pitfalls – is to do without an advisory board and concentrate on creating a stronger ethical culture within the organization, meaning that employees are both trained to assess ethical choices correctly and that the ability to do so is one of the selection criteria for candidates.
Rather than outsourcing ethical responsibility, it becomes part of the fabric of the business, with the whole workforce fluent in the ethics debate and able to act accordingly. The ethics board could help with this and, ultimately, its most important role could be preparing the business to function without it.