Last month we learned that Cambridge Analytica used the personal data of over 50 million people to manipulate democratic elections in the United States, the UK, Indonesia, Thailand and Kenya.
Cambridge Analytica was able to do this, in large part, because of how Facebook collects and makes available the private data of its 2 billion users. The story made acutely visible that the way in which data is collected and how it is used matters tremendously. It can sway the course of history.
While the story of Cambridge Analytica is particularly sensational, it reflects a much broader reality. Today, more data is collected about virtually every part of our lives than ever before. And, increasingly, it is being combined and used by algorithms that can affect the lives of individuals in profound ways.
Sometimes data can be used in ways that have a substantial public benefit. For instance, New York uses data from dozens of databases across the city to inform an algorithm that helps building inspectors identify buildings at high risk for fire. Data can save lives. But it is by no means inevitable that the outcomes of data analytics will be good.
Ensuring the responsible, fair and ethical use of data is one of the most important challenges our society faces today. Technology is evolving very quickly and it is being deployed in many high impact sectors. If we don’t collectively demand that these systems are designed in safe and fair ways, then we risk building the wrong future.
At Enigma we enable some of the world’s largest companies to connect data together so that they can do their jobs more efficiently--whether that’s making sure drugs are manufactured safely or ensuring that financial networks aren’t used for human trafficking.
In all of this work, we strive to use data responsibly and in a manner that’s thoughtful about its broader societal impact. It’s not always easy to do this. Thankfully, there is a rich and challenging set of conversations and analyses that are working to make visible the risks associated with data analytics and to ensure that good norms are established.
Below, we have collected some of the pieces that have been touchstones in our thinking. Data ethics are a complex and emerging issue and we’re sure we’ve left much off the list. Have other ideas? We’d love to hear from you on twitter @enigma_data or via email@example.com.
What we’re reading:
- ProPublica’s Machine Bias series on discriminatory bail algorithms and unfair insurance pricing in minority neighborhoods.
- Facebook Figured Out My Family Secrets, And It Won't Tell Me How by Kashmir Hill for Gizmodo
- “The Trouble With Bias”, Kate Crawford’s keynote presentation at NIPS 2017
- Why Stanford Researchers Tried to Create a ‘Gaydar’ Machine by Heather Murphy for the New York Times
- Machines Taught By Photos Learn a Sexist View of Women by Tom Simonite for Wired
- Principles for Accountable Algorithms from Fairness, Accountability, and Transparency (FAT) in Machine Learning
- Public Scrutiny of Automated Decisions: Early Lessons and Emerging Methods from Upturn and Omidyar Network
- AI Now’s 2017 report and recommendations on the use of AI technologies