Worth Knowing
Intel To Outsource Some Of Its Chip Manufacturing: The incoming CEO of Intel Corporation, the world’s largest chipmaker, confirmed plans to outsource some of its manufacturing to external foundries. The confirmation follows months of speculation about whether the Silicon Valley company, known for designing and manufacturing its own chips, would outsource production, but questions remain about which chips will be outsourced and to which external partners. Some observers are concerned the decision heralds a bleak future for domestic chipmaking. Others have warned that the outsourcing model could make the supply chain more fragile, with only a handful of foreign companies, Taiwan-based Taiwan Semiconductor Manufacturing Company chief among them, responsible for much of the world’s semiconductor manufacturing. Intel had been under some pressure to split its design and manufacturing operations entirely to more closely resemble rivals such as AMD and Nvidia — both of which outsource their manufacturing to chipmakers such as TSMC and Samsung — but incoming CEO Pat Gelsinger said he expected a majority of Intel’s chips would be manufactured in-house in the coming years.
Google Clashes With Another AI Ethics Researcher: A month after its acrimonious split with Timnit Gebru, the former co-lead of its Ethical AI team, Google suspended the corporate access and locked the internal account of Margaret Mitchell, the current head of the team and Gebru’s former co-lead, and launched an investigation into her activities. Axios reported that Mitchell had used an automated program to search her messages for evidence of discriminatory treatment of Gebru before Google locked her account. Mitchell’s suspension is the latest of the company’s internal issues to go public. While Gebru’s ouster garnered the most attention, the formation of the Alphabet Workers Union, reports of Google’s tightening control of internal research on “sensitive topics,” and the alleged firing of a diversity recruiter who criticized the company’s hiring policies have increased scrutiny of the company’s internal policies.
MIT Faculty Defend Colleague Accused Of Hiding Ties To China: Hundreds of MIT faculty members have signed an open letter defending their colleague Gang Chen — a Chinese-born, naturalized U.S. citizen and engineering professor — following his January 14 arrest on charges of grant fraud. While the Justice Department alleges Chen concealed $29 million in foreign funding and other connections to the PRC and PRC-related entities while applying for a Department of Energy grant, his colleagues say the charges mischaracterize “normal academic and research activities.” MIT President L. Rafael Reif issued a statement noting that the funds Chen is accused of concealing were given to MIT as an institution, not to Chen personally, and the university is paying for his legal defense. The arrest came during the final week of the Trump presidency, which took a hard line on issues of Chinese economic espionage as part of its “China Initiative.” While the Biden administration appears likely to maintain a similar stance on intellectual property theft, some observers say arrests like Chen’s could have a chilling effect, discouraging foreign-born students who would otherwise study and work in the United States.
Government Updates
NSCAI Releases Draft Final Report: The National Security Commission on Artificial Intelligence released the draft of its final report in January. The report predicts a near future in which AI will transform all aspects of warfighting, intelligence gathering and espionage. It warns that U.S. competitors, such as China and Russia, may field superior AI-enabled military systems within the next five years, “leapfrogging” the United States unless it changes course. The commission recommends the Pentagon set up a foundation for AI integration by 2025, comprising a DOD-wide digital ecosystem, a technically literate workforce, and more efficient business practices aided by AI. One aspect of the report that received significant public attention was its recommendation that the United States explore autonomous weapons systems and reject any global treaty prohibiting these systems. The independent commission, established by Congress in 2018 and co-chaired by former Deputy Secretary of Defense Robert Work and former Google head Eric Schmidt, is set to deliver its final report to Congress this March.
House Dems Single Out Algorithms In Letters To CEOs: On January 21, two weeks after a violent mob stormed the Capitol building, Congressional Democrats wrote to the CEOs of social media sites Twitter, Facebook and YouTube (and YouTube’s parent company, Google) urging them to alter their algorithmic recommendation systems. The letters, written by Reps. Eshoo and Malinowski and signed by dozens of Democratic representatives, argue that algorithms designed to maximize user engagement have led to political radicalization and violence. While the letters have no legal force behind them, several of their signatories — such as Rep. Cicilline, chairman of the House Judiciary antitrust subcommittee — hold important oversight positions in the House of Representatives. Some observers have noted the pressure may be having an effect: In the week after the letter, Facebook CEO Mark Zuckerberg told investors the site would be permanently removing political groups from its recommendations.
FBI Seeks Machine Learning Tools To Scrape Social Media: Late last month, the Federal Bureau of Investigation released a request for information about the development of machine learning tools for processing data from open source intelligence, primarily mobile messaging applications and social media platforms. The RFI comes from the FBI’s Counterterrorism Advanced Projects Unit, which an attached document says was established to create “specialized capabilities to exploit mobile messaging and social media platforms.” Bloomberg Government, which first reported the RFI, noted the announcement’s proximity to the Capitol attack and the subsequent migration of extremist posting from larger sites to lesser-known platforms. While an RFI is not a solicitation or evidence of a government contract, it is an initial step in that direction. Qualified businesses have until February 9 to submit their information.
In Translation
CSET’s translations of significant foreign language documents on AI
CSET’s translations of significant foreign language documents on AI
Cyberspace Administration of China Regulation: Measures for Cybersecurity Reviews. This regulation, issued by the Cyberspace Administration of China (CAC) in 2020, affects PRC companies that run China’s “critical information infrastructure.” The regulation requires operators of key PRC network platforms to undergo a CAC cybersecurity review before launching products or services that may impact Chinese national security.
PRC Cybersecurity Schools Standards: Notice on the Publication of Management Methods for the Demonstration Project for Building World-Class Cybersecurity Colleges. This document, issued in 2017 by the predecessor office of today’s Cyberspace Administration of China, sets standards for Chinese universities that wish to stand up their own trial cybersecurity schools. In order to fast-track China’s efforts to train cybersecurity talent, the regulations allow undergraduate cybersecurity majors to enroll in these “demonstration” cybersecurity schools without taking the usual graduate school entrance exam and arrange for government and private funding for the new cybersecurity colleges.
If you have a foreign-language document related to security and emerging technologies that you’d like translated into English, CSET may be able to help! Click here for details.
What We’re Reading
Article: Rethinking immigration policies for STEM doctorates, Michael Roach and John Skrentny, Science (January 2021)
Article: Artificial divide: How Europe and America could clash over AI, Ulrike Esther Franke, ECFR (January 2021)
Book Chapter: Artificial Intelligence: Risks and Opportunities for SOF, Paul Scharre (January 2021)
Report: Asymmetric Competition: A Strategy for China & Technology: Actionable Insights for American Leadership, China Strategy Group (Fall 2020)
Book: If Then: How the Simulmatics Corporation Invented the Future, Jill Lepore (2020)
What’s New at CSET
REPORTS
- The Semiconductor Supply Chain: Assessing National Competitiveness by Saif M. Khan, Alexander Mann and Dahlia Peterson
- Securing Semiconductor Supply Chains by Saif M. Khan
- China’s STI Operations: Monitoring Foreign Science and Technology Through Open Sources by William Hannas and Huey-Meei Chang
- Comparing Corporate and University Publication Activity in AI/ML by Simon Rodriguez, Tim Hwang and Rebecca Gelles
- Assessing the Scope of U.S. Visa Restrictions on Chinese Students by Remco Zwetsloot, Emily Weinstein and Ryan Fedasiuk
- Federal Drive with Tom Temin Podcast: Protecting US research from foreign adversaries featuring Melissa Flagg
- Center for a New American Security: Myths and Realities of China’s Military-Civil Fusion Strategy by Lorand Laskai
CSET has a crowd forecasting platform. Sign up as a forecaster, and take a look at some of the predictions so far:
- (New) How much funding will private U.S. tech companies raise between July 1 and December 31, 2021, inclusive?
- (New) How much funding will U.S. tech startups raise between July 1 and December 31, 2021, inclusive?
- (New) What will be the opening price of GameStop stock on Tuesday, February 16?
- Brookings: A recent report about the future of AI governance cited a CSET issue brief by Dewey Murdick, James Dunham and Jennifer Melot, AI Definitions Affect Policymaking.
- Science Magazine: For an article about U.S. research security, Science Magazine reached out to Senior Fellow Melissa Flagg.
- Associated Press: Ben Buchanan, Director of CSET’s CyberAI Project, spoke to the AP about the motivations behind the SolarWinds hack.
- Axios: An article covering the Biden administration’s National Security Council cited several CSET reports: the 2019 report China’s Access to Foreign AI Technology by Wm. C. Hannas and Huey-Meei Chang, Agile Alliances: How the United States and Its Allies Can Deliver a Democratic Way of AI by Andrew Imbrie, Ryan Fedasiuk, Catherine Aiken, Tarun Chhabra (now senior director for technology and national security on the NSC) and Husanjot Chahal, and Open-Source Intelligence for S&T Analysis by Chhabra, Hannas, Dewey Murdick and Anna Puglisi.
Upcoming Events
- February 3: Stanford Artificial Intelligence and Law Society, Global Talent Flows in AI and Their Implications for the United States featuring Remco Zwetsloot
- February 11: CSET, Modeling a Secure Future: Advanced Methods for Managing Risk and Improving Resilience featuring Reginald Brothers
What else is going on? Suggest stories, documents to translate & upcoming events here.