S.F. Passes Facial-Recognition Ban; Capitalizing on AI’s Opportunities

San Francisco passes facial-recognition surveillance ban. San Francisco on May 14 became the first U.S. city to pass a ban on the use of facial recognition by local agencies, reported WSJ’s Asa Fitch. The move comes amid a broad push to regulate the technology, which critics contend perpetuates police bias and provides excessive surveillance powers, although San Francisco’s own police force doesn’t use it.

San Francisco isn’t alone. Similar bans have been proposed in Oakland, Calif., and Somerville, Mass.

Facial recognition proponents cite the technology’s benefits. Law-enforcement groups say bans are an overreaction, adding that the technology can assist in catching criminals and locating missing people when used with police investigative techniques. Dozens of police forces around the country use the technology to analyze mug shots and driver’s-license photos to identify suspects.

Opponents troubled by facial recognition’s possible flaws. Researchers at the Massachusetts Institute of Technology found that facial-recognition tools created by Amazon.com Inc. and others had significantly higher error rates when identifying darker-skinned people and women, according to the WSJ. Amazon disputed the findings. Critics say the system’s flaws raise concerns when the technology is used in decisions that affect people’s liberty.

What’s needed to capitalize on AI’s opportunities. AI has the potential to alter economic growth, commerce and trade. But for AI to develop, there need to be new regulations for AI ethics and data access as well as a revisiting of existing regulations and laws around privacy and intellectual property, according to a report from the Brookings Institute. There also needs to be an international AI development agenda to avoid having a variety of unnecessary regulations that impede the technology’s adoption.

The Brookings Institute offers a number of suggestions for maximizing AI’s benefits, among them:

Strengthen AI diffusion within and across countries. A lack of AI diffusion is producing a widening gap between companies in business sectors. “Policies are needed to increase the rates and depth of technology diffusion across the economy,” according to the report.

· Develop education and skills. AI will require science, technology, engineering and math education, and training. And those skills need to be developed globally.

· Establish sound cybersecurity. Governments need to develop national cybersecurity strategies. However, international cybersecurity standards could create unneeded barriers to trade in AI, the report says.

· Protect the privacy of personal data. Strong rules are needed, but, at the same time, those rules shouldn’t put unnecessary restrictions on cross-border data flows.

· Domestic agenda. A “consistent, transparent, and standardized framework” is needed for sharing data sets across government agencies, researchers and the private sector, the report states.

· Develop a balanced intellectual-property framework. AI technology needs a supportive intellectual-property rule base that includes fair-use exceptions to copyright, which would allow duplicating data for training purposes.

· Develop AI ethical principles. In order to develop trust in artificial intelligence, AI processes and outcomes need to be ethical, according to the report.