Government by algorithm: Difference between revisions

Revision as of 18:15, 4 April 2020

Revision as of 18:17, 4 April 2020

Line 37:

{{Expand section|date=March 2020}}

The are potential risks associated with the use of AI in government. Those include AI becoming susceptible to bias,<ref name=":1">{{Cite web|url=https://ash.harvard.edu/files/ash/files/artificial_intelligence_for_citizen_services.pdf|title=Artificial Intelligence for Citizen Services and Government|last=Mehr|first=Hila|date=August 2017|website=ash.harvard.edu|access-date=2018-12-31}}</ref> a lack of transparency in how an AI application may make decisions,<ref name=":6">{{Cite web|url=https://www.capgemini.com/consulting/wp-content/uploads/sites/30/2017/10/ai-in-public-sector.pdf|title=Unleashing the potential of Artificial Intelligence in the Public Sector|last=Capgemini Consulting|date=2017|website=www.capgemini.com|access-date=2018-12-31}}</ref> and the accountability for any such decisions.<ref name=":6" /> There is also a serious concern that [[gaming the system|gaming]] by the regulated parties might occur, once algorithmic governance brings more transparency is brought into the decision making by algorithmic governance, regulated parties might try to manipulate their outcome in own favor and even use [[adversarial machine learning]].<ref name=sstandford/>

==See also==