ICE drops its pursuit of extreme vetting software
US immigration officials have ended their investigation into the use of AI to determine if visitors might commit criminal acts.
Some background: Last summer, Immigration and Customs Enforcement (ICE) announced their interest in software that could scan the internet to determine if a foreign visitor might pose a threat.
The news: According to the Washington Post, after finding no ready made software to apply or cheaper custom solutions, ICE decided to abandon its hopes of an artificial intelligence system to accomplish this task. It now plans to rely on 180 humans to perform the process instead.
Why it matters: Determining good and bad content (let alone people) using an AI program is extremely difficult. Moving away from this tech for now will ensure AI is not used prematurely for something as high stakes as labeling people as potential terrorists.
Some background: Last summer, Immigration and Customs Enforcement (ICE) announced their interest in software that could scan the internet to determine if a foreign visitor might pose a threat.
The news: According to the Washington Post, after finding no ready made software to apply or cheaper custom solutions, ICE decided to abandon its hopes of an artificial intelligence system to accomplish this task. It now plans to rely on 180 humans to perform the process instead.
Why it matters: Determining good and bad content (let alone people) using an AI program is extremely difficult. Moving away from this tech for now will ensure AI is not used prematurely for something as high stakes as labeling people as potential terrorists.
No comments:
Post a Comment