Artificial Intelligence: How Little Has to Go Wrong

The report outlines risks associated with AI and robotics, including lethal autonomous weapons, accidental nuclear war, and economic disruption. It highlights a survey of scientists and technologists where over 80% believe there is a medium to high chance of AI outcomes going badly wrong and 96% advocate for increased regulation.

• The report identifies several high-risk areas of AI and robotics development, including lethal autonomous weapons, the possibility of accidental nuclear war, the spread of fake news, threats to online privacy, driverless cars, AI in the home, and the potential for economic disruption and mass unemployment.
• There is a concern that without proper regulation, AI and robotics will reinforce existing societal power imbalances and diminish the accountability of governments and corporations.
• A survey of scientists and technologists, primarily members of Scientists for Global Responsibility, revealed that 96% of respondents believe AI requires more regulation, and over 80% perceive a medium to high probability of AI resulting in negative outcomes.

APA

Simms, A., & Parkinson, S. (2018). Artificial Intelligence: How Little Has to Go Wrong. Scientists for Global Responsibility. https://www.sgr.org.uk/publications/artificial-intelligence-how-little-has-go-wrong

Chicago

Simms, Andrew, and Stuart Parkinson. Artificial Intelligence: How Little Has to Go Wrong. Scientists for Global Responsibility, 2018. https://www.sgr.org.uk/publications/artificial-intelligence-how-little-has-go-wrong.

Wikipedia

{{cite report |last1=Simms |first1=Andrew |last2=Parkinson |first2=Stuart |title=Artificial Intelligence: How Little Has to Go Wrong |publisher=Scientists for Global Responsibility |date=November 2018 |url=https://www.sgr.org.uk/publications/artificial-intelligence-how-little-has-go-wrong |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{simms2018artificial, author = {Simms, Andrew and Parkinson, Stuart}, title = {{Artificial Intelligence: How Little Has to Go Wrong}}, institution = {Scientists for Global Responsibility}, year = {2018}, month = nov, url = {https://www.sgr.org.uk/publications/artificial-intelligence-how-little-has-go-wrong}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }