AI R&D and AI-Driven Asymmetric Warfare (ADAW)

The world’s major powers are investing in AI research and development. China has a national AI strategy with a plan to invest $150 billion by 2030. The U.S. DoD invests $7.4 billion a year on unclassified R&D for AI and related fields. Russia, who is currently both light on human capital and financial resources, is currently investing $12.5 million a year.

With a lack of available resources Russia has adjusted its tactical approach to one of asymmetric warfare, where strategic influences and attacks have the potential to produce the greatest impact for their cost. For example, it is estimated that Russia spent about $1 million to influence the 2016 U.S. elections: $100,000 on Facebook ads, $4,700 on Google ads, $240,000 over two years on the IRA troll farm, $50,000 for an intelligence gathering trip by agents in 2014, plus costs for setting up 36,000 automated Twitter bot accounts, costs for creating divisive content, and costs related to the cyber-attacks on the Clinton campaign and the DNC.

The new threat in the information space is AI-driven asymmetric warfare (ADAW). As AI technology advances, the content of disinformation is changing from static (memes, ads, fake news stories) to dynamic (video, audio, "deep-fakes"). AI is also being researched for the purpose of understanding and manipulating human emotions and responses. The manipulative content is able to selectively reach targets via existing advanced content distribution networks.

Summarized from a November 15, 2018 Brookings Report "Weapons of the weak: Russia and AI-driven asymmetric warfare."

Additional resources:
Congressional Artificial Intelligence Caucus
CSIS "Promising Start, but Few Details in House AI Report"

“Digital information warfare is cost-effective and high-impact, making it the perfect weapon of a technologically and economically weak power,” @apolyakova writes on Russian disinformation campaigns https://t.co/8s4qNNpv7v pic.twitter.com/qG9e4hALlh

— Brookings (@BrookingsInst) November 26, 2018

GAO Report on Artifical Intelligence

According to a June 2018 Government Accountability Office report on artificial intelligence, AI can be understood in the waves that the technology evolved. The first wave describes AI via the use of algorithms and includes programs like those that help prepare taxes or schedule logistics. The second wave of AI is based on machine learning and unlike first wave, second wave systems are designed to perceive and learn. Second wave systems include those such as natural language processing and self-driving vehicles. The third wave of AI combines both the first and second waves and adds greater contextual sophistication. An example is a ship at sea that can navigate without human intervention for a few months at time.

The GAO report further evaluates AI’s impact in four key areas: cybersecurity, automated vehicles, criminal justice, and financial services. Each use case raises questions such as those probing regulatory frameworks, data sharing, safety and privacy, ethical considerations, data accuracy, etc.

The report was part of the testimony provided by GAO chief scientist Dr. Tim Persons for a House Science, Space and Technology Subcommittee Hearing “Artificial Intelligence – With Great Power Comes Great Responsibility” in June 2018. Other witnesses included Dr. Jaime Carbonell of Carnegie Mellon University who was absent due to a medical emergency, Mr. Greg Brockman of OpenAI, and Dr. Fei-Fei Li of AI4ALL.