Computers that learn words from texts written by humans capture their meaning but also our biases, a new study shows.

Why it matters: Machine learning is being eyed to sift through resumes in an effort to reduce discrimination in hiring, analyze loan applications and to predict criminal behavior while reducing racial profiling. The unintended biases found in artificial intelligence raise ethical questions about whether and how to deploy the technology without reinforcing stereotypes. (See Exhibit A, the racist Microsoft bot.)

How it works: The researchers created a test for how closely the AI associates different words and uncovered gender and racial biases similar to those of humans that are well-known from psychological studies. They found that European-American names were more closely associated with pleasant words (honest, gentle, happy) whereas unpleasant words (divorce, filth, jail) were more likely to be attributed to African-American names. Young people were considered pleasant, old people were not. They then looked at gender bias and found the AI associated women more so than men with family and the arts than with mathematics.

Thought bubble: Context provides bias but also meaning. How much bias can be removed before that meaning is lost?

The study authors don't recommend untraining the machine because of the risk of removing crucial knowledge about the world. "Artificial intelligence learns biases but it needs the awareness not to make prejudiced decisions. Since machines do possess self-awareness the way humans do, a human in the loop can help machines make ethical decisions," says Princeton's Aylin Caliskan.

Go deeper

Updated 21 mins ago - World

Scoop: Decisive meeting could lead to Israeli-Sudanese normalization

Lt. Gen. Abdel Fattah al-Burhan, Sudan's head of the Sovereign Council, meets with Bahraini aid officials in Khartoum, Sept. 15. Photo: Mazen Mahdi/AFP via Getty Images

U.S., Emirati and Sudanese officials will hold a decisive meeting in Abu Dhabi on Monday on a possible normalization agreement between Sudan and Israel, Sudanese sources told me.

Why it matters: If the U.S. and the United Arab Emirates accommodate Sudan’s requests for economic aid, an announcement on a normalization agreement with Israel similar to the ones struck with the UAE and Bahrain could be made within days, sources briefed on the process tell me.

Updated 2 hours ago - Politics & Policy

Coronavirus dashboard

Illustration: Sarah Grillo/Axios

  1. Global: Total confirmed cases as of 12:45 p.m. ET: 30,838,610 — Total deaths: 958,090— Total recoveries: 21,088,552Map.
  2. U.S.: Total confirmed cases as of 12:45 p.m. ET: 6,777,026 — Total deaths: 199,352 — Total recoveries: 2,577,446 — Total tests: 94,211,463Map.
  3. Politics: Testing czar on Trump's CDC contradictions: "Everybody is right" Trump's health secretary asserts control over all new rules.
  4. Education: What we overlooked in the switch to remote learning
  5. Health: The dwindling chances of eliminating COVID-19.
  6. World: Guatemalan president tests positive for COVID-19 — The countries painting their pandemic recoveries green.
Updated 2 hours ago - Health

7 states set single-day coronavirus case records last week

Data: Compiled by Axios; Map: Danielle Alberti/Axios

Seven states set new highs last week for coronavirus infections recorded in a single day, according to the COVID Tracking Project and state health departments. Wisconsin and Nebraska surpassed records set the previous week.

Why it matters: Problem spots are sticking in the Midwest, although the U.S. is moving in the right direction overall after massive infection spikes this summer.