As the COVID-19 pandemic continues to tear across the globe, we have all become more reliant on the services of the "technology industry" in our everyday lives. However, the pandemic and changes to how we are living have also dramatically affected the way we use, think about and handle technology and intangible data.
From the spread of misinformation on social media to sharing data with governments, some of the key issues are explained and discussed below:
Quick Read:
- Spread of misinformation: Social media sites are a hotbed for the spread of misinformation on COVID-19, as algorithms promote highly engaged with content which is likely to include shocking content relating to the virus. However, social media companies are working with global health organisations to promote verified information on COVID-19 and prevent dangerous information spreading.
- Rise of AI: Content moderation has become a headache for social media sites who cannot allow contractors, who usually undertake moderation, into their offices due to fears of spreading the virus but equally cannot let them work from home due to data privacy considerations. The content moderation void is being filled by full-time employees of these companies and by rolling out artificial intelligence systems to automate content moderation.
- Sharing data with governments: Data on individuals held by private tech-companies could be key to understanding the spread of COVID-19 and the fightback against it. However, private companies are reluctant to share this with governments, even where anonymized, as the long term consequence could be stricter regulations imposed on them.
- Centralised control: China's use of a colour-coded app to assess the condition of its citizens has been effective in the fight back against COVID-19. However, this may have given the Chinese government the pretext to track its citizens, particularly worrying given that Chinese laws governing data privacy are insubstantial.
- Network pressure: Internet infrastructures are being strained as people work from home and turn to streaming services and online gaming for entertainment. Although some streaming services have cut the quality of their service to ease this pressure, British telecom companies are confident that there is enough headroom in the networks for them to remain resilient.
Detailed Read:
1. Spread of misinformation:
Whilst not all information posted on social media is spurious, there is a spread of misinformation on social media sites, whether intended to be malicious or not. This is perpetuated by social media algorithms which promote certain content; shocking posts about COVID-19 (often-click-bait) are likely to receive high user engagement due to the prevailing public uncertainty and apprehension and are thus promoted to other users.
Dangerous content relating to COVID-19, such as posts suggesting drinking bleach can kill the virus, are being actively removed by social media companies. However, posts which are prima facie innocuous remain on these sites regardless of whether they are accurate or not, such as unverified pictures of texts from doctors describing the situations in hospitals.
This is because social media companies are wary of becoming the arbitrator deciding which information is accurate. They faced the same issues in the run-up to the 2016 EU referendum and 2016 US Presidential information; despite the prevalence of "fake news" during these campaigns, social media companies took no action to avoid accusations of manipulating the narrative in favour of a particular campaign.
The current situation is easier for social media companies to navigate as there are centralised institutions providing validated medical advice, such as the World Health Organisation (WHO) and national governments. As such, social media companies can simply promote the "correct narrative" as decided by these bodies. Many sites are providing the WHO free ads on their website and rerouting users to official guidance via COVID-19 hashtags. As of 25 March 2020, Facebook is removing posts relating to conspiracy theories flagged as dangerous by global health organisations whilst YouTube is promoting verified videos providing information on COVID-19 on its homepage. Instagram, Twitter and Facebook are all banning ads which seek to exploit the COVID-19 pandemic.
The UK government's plan to give Ofcom, the UK's media regulator, greater powers to fine social media companies which allow harmful material on their sites, as set out in the "Online Harms White Paper - Initial consultation response" published on 12 February 2020. However, will the social media companies' positive response to the COVID-19 soften the government's regulatory approach?
2. Rise of AI:
Despite social media companies' best intentions to prevent the spread of misinformation on their sites, they may not have the capacity to stop it due to a deficit of content moderators.
Normally, these companies employ thousands of contractors to moderate content on their sites. For privacy and security reasons, these contractors work on designated production floors in offices; they must scan in and out and must not use personal devices on these floors.
COVID-19 creates a problem for social media companies. Keeping these production floors open significantly increases the likelihood of the virus spreading and is contrary to official guidance. However, letting contractors, who access personal data on a daily basis, work from home risks a huge data privacy breach or, even if there was no actual privacy breach, a PR disaster.
Initially, Facebook and other social media companies kept their production floors open, despite recommending their full-time employees work from home. However, after criticism, they have now closed their production floors as well. Does this initial difference in approach raise employment law questions? Although contractors have different terms of employment to full-time employees, asking contractors to potentially risk their lives by the mere fact of their differing employment status raises tricky legal issues.
So who is moderating the content on their sites relating to COVID-19 (as well as the usual content)? Some moderating duties have been shifted to full-time employees. However, these tech companies have also started automating content moderation through artificial intelligence programmes. Although this was the ultimate goal of social media sites anyway, they have had to accelerate the roll-out of machine learning systems to cope with the COVID-19 crisis.
Early indications suggest the automated content moderation is strewn with errors. However, as bugs are found and fixed, this will improve. The question is whether this will happen quickly enough to stop the spread of misinformation on COVID-19.
3. Sharing data with governments:
As the COVID-19 pandemic worsens, more pressure is piling on technology companies to deploy their resources to assist governments in combatting the virus.
In some instances this is unproblematic. In the US, IBM is using AI (such as its Watson Natural language processing) to turn written text on COVID-19 into machine-friendly data and pairing this information with its weather channel app to track COVID-19 cases at a county and state level. Further, IBM, Google, Amazon and Microsoft are providing COVID-19 researchers access to supercomputers to aid their research.
However, the key to understanding how to beat the COVID-19 is accessing data on millions of individuals, thereby allowing scientists to assess factors that might affect transmission rates and identify trends. SparkBeyond has deployed its AI-driven analytics to collect publicly available data on confirmed cases in Italy, and their subsequent journeys, and built a map predicting the likelihood of contracting the virus in a particular area.
Yet, publicly available data alone is unlikely to be enough. The key to turning the tide may be location data accessed through individual mobile phones. Public health experts could request private-sector companies compile anonymous, aggregated data which could be used to track the spread of the virus as well as assess public behaviour and whether social distancing measures are being complied with. In the UK, mobile network O2 are already working with the government to analyse anonymous smartphone location data to see whether people are adhering to social distancing guidelines.
This creates a dilemma for tech companies like Google, Facebook and Apple. Although providing this data to governments would greatly assist in the fight against COVID-19, it would also reveal the true scope of these companies' abilities to track and hold data on their customers. No matter how worthy the cause currently is, the long term consequences of divulging the extent of information held on customers could be tighter restrictions on tech companies, especially given the recent scandals relating to the misuse of customer data such as, inter alia¸ the Cambridge Analytica scandal. Thus, Google has stated it has not shared anonymized public data with the government and has no plans to do so; will the rest of the industry dig in and take the same approach?
If they do, it is unclear whether there is any legal basis (both in the UK and US) allowing the respective governments to compel these private companies to hand over private information on customers. Even if a sound legal argument could be formulated, the negative optics of using the law to ride roughshod over private companies' and citizens' rights would likely prevent the governments from pursuing this course, no matter how worthy the cause.
4. Centralised control:
China, the first country to be affected by COVID-19, is now recovering and life is starting to return to normal. This has been credited in part to the stringent measures imposed by the Chinese government in order to fight the virus.
Chinese companies have been installing facial recognition technology which can detect individuals in a crowd who have an elevated temperature or who are not wearing a facemask. Chinese citizens have also been required to download an app (run through a platform owned by Alibaba) which assigns them a QR colour code: green means they can move around unrestricted; yellow means a 7-day quarantine; red means a 14-day quarantine. Chinese citizens need to scan this QR code to access public buildings and transport as well as their own apartment blocks. A new feature is being developed which allows other citizens to check an individual's "health status".
These measures seemed to have worked with infection rates falling significantly across the country. However, at what cost? The New York Times analysed the app's code and found it sends a person's location and identifying code to the Chinese authorities every time the QR code is scanned, allowing a person's movements to be tracked. The COVID-19 crisis has given the Chinese government the pretext to track its citizens, particularly worrying given that Chinese laws governing data privacy are insubstantial. Such powers are also unlikely to be rolled back when the COVID-19 pandemic is over given the precedent set by the installation of mass surveillance for the 2008 Olympic Games and 2010 Shanghai Expo which became permanent after the events ended.
Yet, given the behaviour of the British pubic during the COVID-19 crisis, should a similar app system be implemented in the UK? Although the long term monitoring of citizens by governments is clearly dangerous, at a time of crisis should individuals' rights be impinged to prevent the spread of the virus? Given the stretched resources of the emergency services, does utilising data to ensure British citizens are adhering to social distancing measures sound that bad if it would save lives? However, even if the UK government wanted to roll out such an app, they would need to consider whether it was compatible with the General Data Protection Regulation (GDPR) and the Data Protection Act 2018 which protects citizens' data and privacy rights. Any attempt to roll out such an app would almost certainly face a judicial review.
5. Network pressure:
With countries all over the world on lockdown, the surge of people working from home or turning to HD streaming and online gaming for entertainment is straining global internet infrastructure. A heat map created by KASPR Datahaus, using measurements on the differences in the speed of internet packets travelling between the US and destinations regions, shows the pressure on local networks.
The EU commission has reached agreements with streaming services such as YouTube and Netflix to reduce streaming quality, thereby easing pressure on broadband networks. This has led some Netflix customers asking for a 25% reduction in their subscription price to reflect the 25% decrease in streaming quality implemented by Netflix. Disney has also delayed the launch of Disney+ in France by two weeks at the behest of the French authorities due to concerns over the increased traffic it will cause.
Do customers who pay subscriptions to such sites have claims against their providers if the quality of the service provided is reduced? It is likely these companies' terms and conditions allow them to reduce the streaming quality but customers could still have a claim pursuant to the Supply of Goods and Services Act 1982. However, the merits of such a claim are dubious, especially given the context of the decrease in streaming quality.
Ofcom has launched the "Stay Connected" campaign in which the largest British telecom companies, such as O2, Sky, Virgin Media and BT have joined forces to manage the strain on internet and mobile services. The campaign suggests using landlines where possible and turning Wi-Fi off on devices not being used.
Yet, these companies remain sanguine about the resilience of the broadband networks due to the headroom built into them. BT Open-reach has reported a 20% increase in daytime usage across their fibre network since the COVID-19 pandemic began. However, this is not as high as usage levels during peak evening hours in normal circumstances and therefore the networks are well within their capacity limits.