When Ren Zhengfei founded Huawei in 1987, the Chinese telecommunications company had a few thousand dollars in the bank and an eye on reverse-engineering advanced foreign technology. By 1994, Huawei was producing switching equipment, the hardware and software that form the basis of modern telecommunications, and Ren was taking a meeting with Chinese Communist Party Secretary General Jiang Zemin. The Huawei chief executive observed that his company’s product was a matter of “national security” and that a country “that did not have its own switching equipment was like one that lacked its own military.” “Well said,” Jiang replied. From that moment on, business and government have been partners in the mission to keep China’s telecommunications secure.

Meanwhile, in the late 1990s and throughout the first decade of this century, the United States did not think much about its telecommunications dominance. Its leadership in the sector was unrivaled, and U.S. innovations—including 2G, 3G, and 4G technologies—were widely adopted and securely used around the world. But as the United States sat complacently in the top position, trusting that the strength of the free market would keep it there, China was carefully setting itself up as a challenger. Beijing poured resources into Huawei and other Chinese companies, positioning them to outcompete foreign firms. So successful was this effort that by 2012, Huawei telecom gear had been deployed across rural America, covering the bases that house U.S. nuclear weapons. This effectively provided the Chinese government with constant surveillance of the United States’ most sensitive capabilities and military operations. Huawei may not have been raking in profits from this venture, but for Beijing it was an intelligence coup.

Continue reading at foreignaffairs.com

In the midst of Elon Musk’s Department of Government Efficiency’s headline-making actions, it’s a good time to remember that DOGE’s mandate is to “maximize governmental efficiency and productivity” by “modernizing federal technology and software.” Federal IT is a worthy target for Musk and his team of efficiency hackers. It’s overpriced, with various estimates pegging the government’s costs at two to four times the private sector’s. It underdelivers compared to what’s available outside of government, as anyone who has held positions in both sectors will attest. And it is difficult to defend against cyber-attacks — not least because IT vendors upcharge for security features against flaws in their own products.

I contributed to two efforts that chipped away at the cybersecurity aspects of the problem: President Obama’s Cybersecurity National Action Plan and President Trump’s executive order on cybersecurity. After I left the White House in 2017, the first Trump administration and the Biden administration both made additional, incremental progress, especially in migrating agencies to the cloud. And yet federal IT has proven remarkably resistant to reform while also facing the most challenging cyber threat environment of any organization in the world. If DOGE is looking for the most difficult problem in IT, this is it.

Continue reading at federalnewsnetwork.com

In the early 1950s, the United States faced a critical intelligence challenge in its burgeoning competition with the Soviet Union. Outdated German reconnaissance photos from World War II could no longer provide sufficient intelligence about Soviet military capabilities, and existing U.S. surveillance capabilities were no longer able to penetrate the Soviet Union’s closed airspace. This deficiency spurred an audacious moonshot initiative: the development of the U-2 reconnaissance aircraft. In only a few years, U-2 missions were delivering vital intelligence, capturing images of Soviet missile installations in Cuba and bringing near-real-time insights from behind the Iron Curtain to the Oval Office.

Today, the United States stands at a similar juncture. Competition between Washington and its rivals over the future of the global order is intensifying, and now, much as in the early 1950s, the United States must take advantage of its world-class private sector and ample capacity for innovation to outcompete its adversaries. The U.S. intelligence community must harness the country’s sources of strength to deliver insights to policymakers at the speed of today’s world. The integration of artificial intelligence, particularly through large language models, offers groundbreaking opportunities to improve intelligence operations and analysis, enabling the delivery of faster and more relevant support to decisionmakers. This technological revolution comes with significant downsides, however, especially as adversaries exploit similar advancements to uncover and counter U.S. intelligence operations. With an AI race underway, the United States must challenge itself to be first—first to benefit from AI, first to protect itself from enemies who might use the technology for ill, and first to use AI in line with the laws and values of a democracy.

Continue reading at foreignaffairs.com.

"How far away was Oswald from Kennedy?" "Major depressive disorder." 

Thomas Matthew Crooks, the individual who attempted to assassinate former president Donald Trump earlier this summer, made these online searches just before he shot at the former president. He also searched for images of prior mass shooters, the location of the Trump rally where he eventually shot Donald Trump, and the location of a local gun store where he purchased bullets the day of the attempted assassination. Google search has existed for over 20 years, yet it does not recognize dangerous thought processes and respond to the user appropriately.

Continue reading at statnews.com

Key Takeaways

• The rapid advancement and proliferation of AI systems, including foundation models, has catalyzed the widespread adoption of AI benchmarks—yet only very limited research to date has evaluated the quality of AI benchmarks in a structured manner.
• We reviewed benchmarking literature and interviewed expert stakeholders to define what makes a high-quality benchmark, and developed a novel assessment framework for evaluating AI benchmarks based on 46 criteria across five benchmark life-cycle phases.
• In scoring 24 AI benchmarks, we found large quality differences between them, including those widely relied on by developers and policymakers. Most benchmarks are highest quality at the design stage and lowest quality at the implementation stage.
• Policymakers should encourage developers, companies, civil society groups, and government organizations to articulate benchmark quality when conducting or relying on AI model evaluations and consult best practices for minimum quality assurance. 

Read the full brief here

There is an increasing interest in using language models (LMs) for automated decision-making, with multiple countries actively testing LMs to aid in military crisis decision-making. To scrutinize relying on LM decision-making in high-stakes settings, we examine the inconsistency of responses in a crisis simulation ("wargame"), similar to reported tests conducted by the US military. Prior work illustrated escalatory tendencies and varying levels of aggression among LMs but were constrained to simulations with pre-defined actions. This was due to the challenges associated with quantitatively measuring semantic differences and evaluating natural language decision-making without relying on pre-defined actions. In this work, we query LMs for free form responses and use a metric based on BERTScore to measure response inconsistency quantitatively. Leveraging the benefits of BERTScore, we show that the inconsistency metric is robust to linguistic variations that preserve semantic meaning in a question-answering setting across text lengths. We show that all five tested LMs exhibit levels of inconsistency that indicate semantic differences, even when adjusting the wargame setting, anonymizing involved conflict countries, or adjusting the sampling temperature parameter T. Further qualitative evaluation shows that models recommend courses of action that share few to no similarities. We also study the impact of different prompt sensitivity variations on inconsistency at temperature T=0. We find that inconsistency due to semantically equivalent prompt variations can exceed response inconsistency from temperature sampling for most studied models across different levels of ablations. Given the high-stakes nature of military deployment, we recommend further consideration be taken before using LMs to inform military decisions or other cases of high-stakes decision-making.

Find the full paper here.

Abstract

With developments in artificial intelligence (AI) widely framed as security concern in both military and civilian realms, governments have turned their attention to regulating and governing AI. In a study of United States (US), Chinese, and European Union (EU) AI documents, we go beyond instrumental understandings of AI as a technological capability, which serves states’ self-interests and the maintenance of their (supra)national security. Our specific interest lies in how AI policies tap into both problem-solving approaches and affective registers to achieve both physical and ontological securities. We find that in governmental visions, AI is perceived as a capability that enhances societal and geopolitical interests while its risks are framed as manageable. This echoes strands within human–computer interaction that draw on human-centered perceptions of technology and assumptions about human–AI relationships of trust. Despite different cultural and institutional settings, the visions of future AI development are shaped by this (shared) understanding of human–AI interaction, offering common ground in the navigation of innovation policies.

Find the full article at journals.sagepub.com

Windows users are still rightly upset about the global outage in July 2024 that downed 8.5 million Microsoft devices, grounded air travel and disrupted businesses around the world. Although CrowdStrike’s system-crippling update sparked the outage, Microsoft’s antiquated operating system infrastructure was the tinder that led to a global technological wildfire.

In response, Microsoft hosted a summit with CrowdStrike and other endpoint security companies to discuss ways to better protect critical infrastructure and improve security resiliency. The solutions coming out of it are promising, but a central issue remains: interoperability between Windows and third-party security products.

Continue reading at infosecurity-magazine.com

While millions of lives have been saved through medical drugs, many thousands died during the 19th century by ingesting unsafe medicines sold by charlatans. Across the US and Europe this led to the gradual implementation of food and drug safety laws and institutes — including the US Food and Drug Administration — to ensure that the benefits outweigh the harms. 

The rise of artificial intelligence large language models such as GPT-4 is turbocharging industries to make everything from scientific innovation to education to film-making easier and more efficient. But alongside enormous benefits, these technologies can also create severe national security risks.  

We wouldn’t allow a new drug to be sold without thorough testing for safety and efficacy, so why should AI be any different? Creating a “Food and Drug Administration for AI” may be a blunt metaphor, as the AI Now Institute has written, but it is time for governments to mandate AI safety testing.

Continue reading at ft.com