Digital human rights are next frontier for fund groups
Investors take a critical look at potentially damaging role of some technology in everyday lives
Politicians publicly grilling technology chiefs such as Facebook’s Mark Zuckerberg is all too familiar for investors.
“There isn’t a day that goes by where you don’t see one of the tech companies talking to Congress or being highlighted for some kind of controversy,” says Lauren Compere, director of shareholder engagement at Boston Common Asset Management, a $2.4bn fund group that invests heavily in tech stocks.
Fallout from the Cambridge Analytica scandal that engulfed Facebook was a wake-up call for investors such as Boston Common, underlining the damaging social effects of digital technology if left unchecked. “These are the red flags coming up for us again and again,” says Ms Compere.
Digital human rights are fast becoming the latest front in the debate around fund managers’ ethical investments efforts. Fund managers have come under pressure in recent years to divest from companies that can harm human rights — from gun manufacturers or retailers to operators of private prisons. The focus is now switching to the less tangible but equally serious human rights risks lurking in fund managers’ technology holdings.
Attention on technology groups began with concerns around data privacy, but emerging focal points are targeted advertising and how companies deal with online extremism.
Following a terrorist attack in New Zealand this year where the shooter posted video footage of the incident online, investors managing assets of more than NZ$90bn (US$57bn) urged Facebook, Twitter and Alphabet, Google’s parent company, to take more action in dealing with violent or extremist content published on their platforms.
The Investor Alliance for Human Rights is currently co-ordinating a global engagement effort with Alphabet over the governance of its artificial intelligence technology, data privacy and online extremism.
Investor engagement on the topic of digital human rights is in its infancy. One roadblock for investors has been the difficulty they face in detecting and measuring what the actual risks are. “Most investors do not have a very good understanding of the implications of all of the issues in the digital space and don’t have sufficient research and tools to properly assess them — and that goes for companies too,” said Ms Compere.
One rare resource available is the Ranking Digital Rights Corporate Accountability Index, established in 2015, which rates tech companies based on a range of metrics. The development of such tools gives investors more information on the risk associated with technological advancements, enabling them to hold companies to account when they identify risks and questionable ethics.
Another challenge is the pace of technological advancements. “The concern is that technological changes are being quickly driven through by companies but sometimes we don’t stop and ask what is the human rights impact,” said John Howchin, secretary-general of the Swedish AP funds’ ethics council. “In future it will be seen as a wild west.”
This point is highlighted by the fierce debate surrounding artificial intelligence and facial recognition software. Hailed as a way to make society more efficient, investors are grappling with the fact that facial recognition can also be used for unethical purposes. Critics of the software say companies that manufacture and sell it may not be aware of how it is being used, especially when adopted by governments for military purposes.
Facial and emotion recognition systems are notably deployed by the Chinese government for surveillance in the Xinjiang region, where an estimated 1m mostly Muslim minorities are held in internment camps. In the US, activists and privacy groups claim the technology is used in a harmful way by federal agencies such as Immigration and Customs Enforcement and Customs and Border Protection.
There are also concerns about flaws and biases in facial recognition programmes. A test by the American Civil Liberties Union last year found that Amazon’s facial recognition technology, Rekognition, wrongly identified 28 members of Congress as arrestees. Research has also shown that the software registers more cases of mistaken identity with women and minorities.
Carola van Lamoen, head of active ownership at Robeco, the €199bn asset manager, said: “We see the ability of AI to facilitate growth potential for technology companies but we also see the associated risks.”
Complicating the picture is the absence of harmonised regulations governing AI. Robeco recently embarked on a probe into the social impact of AI prompted by concerns that the advancements were outpacing the development of rules or principles.
In the US, for example, a handful of cities have banned local government and law enforcement agencies from deploying facial recognition software but federal agencies can still use it.
“What worries investors [about AI] is that the technology is being developed in a vacuum — we’re not seeing a global norm around it,” said Ms Compere.
Several companies, including Google, Microsoft and IBM, have taken initial steps to protect against unintended risks by adopting codes of AI ethics.
Microsoft has been particularly vocal, calling last year for “a government initiative to regulate the proper use of facial recognition technology”. This year the company declined a request from a California law enforcement agency to install facial recognition technology in officers’ cars and body cameras over human rights concerns.
Despite Microsoft’s decision, few companies have gone public on how risks to human rights influence their business decisions and affect which contracts they accept or decline.
Some investors are using their voice within companies to make a stand. The Interfaith Center on Corporate Responsibility, a coalition of responsible investors managing more than $400bn in assets, this year filed a resolution at Amazon calling for it to stop selling Rekognition to governments until it had completed a review into whether it posed a threat to civil and human rights.
Other investors want companies to integrate human rights considerations into their formal risk management assessments and make senior executives accountable for this. “We want companies to do their homework,” said Ms van Lamoen. “They should know when and how AI is used, assess the potential impacts and actively monitor this at board level.”
“Technology companies have had issues involving human rights over the past few years,” said Mr Howchin. “We think there’s a need for the whole technology sector to better understand human rights and start hiring people who are specialist in human rights.”
This involves asking companies to train their AI designers to consider human rights during the product development process. “Unless concepts are baked in to the actual design of newer technologies from the get-go, it is a lot harder to control human rights implications,” said Ms Compere.