Intelligence chief warns of threats from AI

Getty Images

Advances in artificial intelligence (AI) create new security risks for the United States, a top intelligence official is expected to tell Congress on Tuesday.

“Implications of broader Al deployment include increased vulnerability to cyberattack, difficulty in ascertaining attribution, facilitation of advances in foreign weapon and intelligence systems, the risk of accidents and related liability issues, and unemployment,” Director of National Intelligence James Clapper will say at a Senate Select Committee on Intelligence hearing, according to prepared remarks.

Clapper will also say that “Al systems are susceptible to a range of disruptive and deceptive tactics that might be difficult to anticipate or quickly understand” and will cite instances in which the use of artificial intelligence has caused fluctuations in the stock market.

Growth in the artificial intelligence sector — a focus of companies such as Google and Apple — has produced anxiety in tech circles. Tesla founder Elon Musk and physicist Stephen Hawking signed an open letter last year raising concerns about what could happen when weapons developers use artificial intelligence.

“Starting a military AI arms race is a bad idea, and should be prevented by a ban on offensive autonomous weapons beyond meaningful human control,” that letter says.

Clapper's AI concerns are part of a larger annual report on global threats that also covers other technology-related concerns.

“The consequences of innovation and increased reliance on information technology in the next few years on both our society's way of life in general and how we in the Intelligence Community specifically perform our mission will probably be far greater in scope and impact than ever,” the report says.

Clappper will also single out the potential for vulnerabilities in the connected devices, such as appliances and home technology, as well as self-driving cars. He will note foreign advancements in data science and other countries’ efforts to mine U.S. published data to their advantage as well.