Free beta access. 5 free AI tokens each week. Optional top-ups available. See what is included
Opening Word document…

Focusing on “winning” the arms race will undermine effective oversight and testing – we will rush into AI development without safeguards against their abuse.

Bob, 2019 - intelligence, terrorism and legal analyst at the Jerusalem Post [Yonah Jeremy, 13 February Jerusalem Post, “Top cybersecurity expert: Crisis worse than 9/11 could emerge from AI arms race.” https://www.proquest.com/docview/2182568907?parentSessionId=BxSQxuLDub9YJx%2BOeIXcOzvsMV848Jfpxd5%2B%2B3GON8c%3D&pq-origsite=primo&accountid=14667, 6/18/22 MD]

Winning the artificial intelligence arms (AI) race between the US, Israel and their adversaries may need to take precedent over balancing the risks at this stage, a former government agency chief technology officer has told The Jerusalem Post. Those risks could include crises even worse than 9/11, said Amit Meltzer, now a top cybersecurity consultant. Discussing the issue on Tuesday, merely a day after US President Donald Trump issued the first-ever executive order to bolster US efforts in the artificial intelligence arms race, he said that winning such a competition would not go well with careful oversight of negative consequences and potential abuses. While US Senate Vice Intelligence Committee Chairman Mark Warner (D-VA) applauded aspects of the order, he criticized the lack of developing oversight for the misuse of AI. Warner supported the order’s provisions for opening US federal data sets to non-federal entities to enhance cooperation, but said that Trump’s order 'reflects a laissez-faire approach to AI development that... will have the US repeating the mistakes it has made in treating digital technologies as inherently positive forces, with insufficient consideration paid to their misapplication.' Meltzer called Trump’s declaration more symbolic rather than concretely operational, but said that it still was crucial 'to give strong backing to academic and government institutions… which suffer from a chronic lack of personnel' and resources. Furthermore, he said that concerns that certain companies would quickly gain domination of the AI sector and abuse their standing economically were possible. But, he said that it was nearly impossible to square 'the national necessity for the US' or Israel to 'strengthen and maintain leadership in the industry' with policies that encourage caution and that new technologies should only be rolled out after any danger was carefully examined. He compared the need to act with trying to keep up in the field of artificial intelligence with the US’s sale of weapons to any dictatorship that is not in a direct conflict with it, in order to make weapon development economically sustainable. However, Meltzer said that neither Trump nor Warner honed-in on the true potential dangers of AI. Once AI starts taking over a large number of societal functions, damage to a such a network could have far more disastrous and broader consequences and at a more rapid pace, he explained. By raising the specter of a disaster the size of the stock market crash of 1987, he said that economic or social-psychological warfare influence campaigns that dwarf current threats would become possible. Another nightmare scenario could be using an AI algorithm to make four million Toyota cars all crash at the same time worldwide, leaving countless dead and wounded in a tragedy 'that would be worse than 9/11.' The worst part of these dangers is that artificial intelligence could make such attacks infinitely easier than with advanced existing cyber hacking capabilities. Despite these threats, he predicted that the US needs to keep up with China, Russia and Israel’s need to stay ahead of its adversaries, meaning that rapidly developing new capabilities would continue to be put before protecting civil liberties. 'Afterwards, they will fix things and ask forgiveness from the victims,' he said of moving ahead too fast without proper oversight.

The Rhetoric of an AI Arms Race sparks a race to the bottom – states will deploy unsafe weapons to get them out before their competitors

Scharre, 2019 - Vice President and Director of Studies at CNAS [Paul, May-June, “Killer Apps: The Real Dangers of an AI Arms Race,” https://omnilogos.com/killer-apps-real-dangers-of-ai-arms-race/6/18/22 MD]

The nation that leads in the development of artificial intelligence will, Russian President Vladimir Putin proclaimed in 2017, "become the ruler of the world." That view has become commonplace in global capitals. Already, more than a dozen governments have announced national AI initiatives. In 2017, set a goal of becoming the global leader in AI by 2030. Earlier this year, the White House released the American AI Initiative, and the U.S. Department of Defense rolled out an AI strategy. But the emerging narrative of an "AI " reflects a mistaken view of the risks from AI--and introduces significant new risks as a result. For each country, the real danger is not that it will fall behind its competitors in AI but that the perception of a race will prompt everyone to rush to deploy unsafe AI systems. In their desire to win, countries risk endangering themselves just as much as their opponents. AI promises to bring both enormous benefits, in everything from health care to transportation, and huge risks. But those risks aren't something out of science fiction; there's no need to fear a robot uprising. The real threat will come from humans. Right now, AI systems are powerful but unreliable. Many of them are vulnerable to sophisticated attacks or fail when used outside the environment in which they were trained. Governments want their systems to work properly, but competition brings pressure to cut corners. Even if other countries aren't on the brink of major AI breakthroughs, the perception that they're rushing ahead could push others to do the same. And if a government deployed an untested AI weapons system or relied on a faulty AI system to launch cyberattacks, the result could be disaster for everyone involved. Policymakers should learn from the history of computer networks and make security a leading factor in AI design from the beginning. They should also ratchet down the rhetoric about an AI arms race and look for opportunities to cooperate with other countries to reduce the risks from AI. A race to the bottom on AI safety is a race no one would win.

Competitive races pressure militaries to deploy AI before it is safe, increasing the risk of catastrophic accidents

Arnold and Toner, 2021 – Center for Security and Emerging Threats [Zachary and Helen, July, CSET Policy Brief. “AI Accidents: An Emerging Threat What Could Happen and What to Do” https://cset. georgetown.edu/wp-content/uploads/CSET-AI-Accidents-An-Emerging-Threat.pdf Acc 6/7/22 TA]

As AI is integrated into more and more critical systems, the dangers of AI accidents will grow. In practice, we expect these accidents will be more likely and more severe in some situations than in others. Identifying these risky situations ahead of time is challenging, but based on AI accidents that have already occurred and historical accidents involving other technologies, we expect risk factors for severe AI accidents will include: ● Competitive pressure. When not using AI could mean falling behind competitors or losing profits, companies, militaries, and governments are more likely to deploy buggy AI systems, use them in reckless ways, or cut corners on testing and operator training.39 The infamous Boeing 737 MAX, though it does not use machine learning, is an example of this dynamic. The aircraft was developed, tested, and certified under extreme time pressure, aiming to compete with a comparable Airbus system.40 Ultimately, this haste led to two crashed planes and hundreds of deaths.

Sign in to recut, update, or save this card.