However, there are also great hopes invested in the use of big data for social good. Information about everyday patterns of human behaviour can help us construct algorithms to help prioritise interventions in areas such as education, health or sanitation.
Indeed, big data is central to the so-called data revolution for the Sustainable Development Goals. This revolution is about improving the data that’s used for designing, monitoring and evaluating development interventions, and for holding decision-makers to account.
It’s also about preserving privacy, developing better tools to measure impact, and disaggregate information to ensure that international development leaves no-one behind.
A training session for young sex workers in Myanmar. ©Arkar Kyaw 2016/ International HIV/AIDS Alliance
Big data and the global HIV response
The global HIV response has for decades been using large quantities of data to build economic and epidemiological models to decide smart HIV investments.
In epidemiology, statistical data is used to establish the relationships between individual patterns of behaviour and the spread of disease. Mathematical modelling, also relying on statistical data, is used to simulate the relationship between various possible investment scenarios and an optimum combination of HIV prevention and treatment interventions. This in turn is used as a basis to determine the nature of global HIV investments.
The tyranny of experts
As the production of big data escalates, its potential applications expand. For example, the analysis of hundreds of thousands online postings or web searches about HIV prevention and testing can help identify who to offer HIV self-testing kits or PrEP.
But as Bruno Lepri and others point out, this ’modelisation’ could fall into the so-called ‘tyranny of experts’. This is when only a few experts have access to data and are able to interpret, or at worst, manipulate it. Indeed, the use of modelling for decision-making in the HIV response often lacks transparency and accountability about how models are designed and how the information is stored and treated.
With big data, there is the additional risk of a ‘tyranny of data’. If the data is biased or incomplete, it can perpetuate, and even justify, prevailing patterns of exclusion and discrimination of key populations.
Inaccurate data (e.g. wrong size estimates for key populations) and the absence of critical social and political variables (such as the impact of human rights interventions) in HIV modelling can have disastrous consequences. It can distort priorities and lead to wrong or inadequate responses. This risk is particularly high among key populations in countries where they are stigmatised, excluded and persecuted.
Access to HIV services still lacking
The wider use of big data in the HIV response takes place in a context of continuing lack of access to HIV prevention and treatment services in a growing number of countries. This impending crisis is a breeding ground for social mobilisation and campaigning.
This context is similar to the denial of free AIDS treatment in the 1990s. At that time, people living with HIV had to learn medicine and drug regime development at record speed in order to challenge the tyranny of experts, which was denying them access to experimental, but potentially life-saving drugs.
HIV campaigners took to the streets, to mass media and to courtrooms. They created the biggest global health movement in history.
The new HIV activism
Today’s HIV activism will need to adapt to the ways of the 21st century.
This new mobilisation, we believe, will be among millennials, young people who are the population most affected by HIV at present.
This time, we think, the HIV revolution will rise from the South, not the North. It will be unstructured and decentralised. It will be online, and will happen through mobile technologies.
The new HIV activism will need to break the tyranny of data. We will need to understand big data better; to benefit from it, but also to challenge data management methods that suggest discriminatory investments or breach the privacy, security or other rights of people affected by HIV.
We will also need to break the tyranny of experts. We will need to reclaim our full and meaningful involvement in the processes that design and use big data to inform HIV policies and investment decisions.
Big data challenges and opportunities are for us both great and inevitable.