As part of an effort to more closely examine the relationship between academia and industry at Stanford, The Daily looked into Stanford’s Industrial Affiliates program, which facilitates a direct connection between university research and private corporations.
Though the specifics of the programs vary by academic department, each of the 68 participating departments allow “companies [to] receive facilitated access to research programs and to participating faculty and students” in exchange for variable “corporate membership fees,” according to the University Corporate and Foundation Relations website. Members of all industrial affiliate programs “attend annual meetings, receive copies of reports and publications, and have opportunities to recruit students.”
Annual membership fees vary among departments, ranging from $15,000 for the Statistics Industrial Affiliates Program to $500,000 for the Premium Stanford Platform Laboratory Affiliates Program.
The programs span an array of fields and are often interdisciplinary in nature, ranging from the Center on Longevity and the Center for the Advancement of Women’s Leadership to Biopharmaceutical Sciences and Thermal and Fluid Sciences. Accordingly, the companies involved also span various fields, such as transportation (Chevron, Boeing), technology (Apple, Baidu), healthcare (Omnicell) and construction (Mortenson).
Affiliation with internal companies
While most participating companies are U.S.-based, there are a significant number of international ones, many of them situated in China, which has presented some issues within the University. One such issue is related to Stanford’s moratorium on research funding from Huawei, a Chinese telecommunications company currently on the Department of Commerce’s Entity List, which identifies foreign businesses subject to specific license requirements for the export and transfer of specified technologies. Huawei was a member of many Industrial Affiliates, including the Stanford Center for Image Systems Engineering (SCIEN) before the moratorium was put in place.
In recent months, federal agencies like the Department of Energy and National Institutes of Health have implemented new regulations against “undue foreign influence,” aimed to protect U.S. intellectual property from threats abroad, primarily China. According to SCIEN director Joyce Farell, professor and Vice Provost and Dean of Research Kam Moler told faculty in an October meeting that the moratorium was put in place to allow Stanford to develop a clear set of policies to ensure that faculty collaboration with Huawei did not violate any federal regulations, which could potentially place university federal funding at risk.
Were it not for the moratorium, approximately $6-7 million in gift donations from Huawei meant to support students would have been granted, according to data compiled by affected faculty.
In late October of this year, Moler sent a letter to faculty detailing six actionable items dealing with foreign research relationships, two of which are to “review industry affiliates programs” and “establish a foreign engagement review program” that provides advice on “foreign influence risk.” In light of increased federal concern over collaboration with foreign companies, members of the Stanford community have expressed concerns about ensuring a balance between national security and the exchange of knowledge with foreign countries.
One program that collaborates extensively with global corporations is the Artificial Intelligence (AI) Affiliates program, which includes corporations such as Google, Chinese transportation company DiDi and South Korean telecommunications company SK Telecom. The program requires partner corporations to “provide $200,000 per year of unrestricted support with an expected three-year commitment,” according to its website. In exchange, corporations can recruit students, access AI lab research results, and attend all retreats, conferences and seminars, per their corporate overview. For additional charges, companies can also establish a visiting researcher at Stanford.
The Daily reached out to the 18 faculty members of the Stanford Artificial Intelligence Lab — a leading hub of artificial intelligence research, teaching, theory and practice — but the faculty were all either unaware of the program, didn’t respond or declined to comment on it.
All Industrial Affiliates programs must abide by Stanford’s establishment policies. The policies, which are intended to “preserve the University’s academic integrity and independence,” include: (1) compliance with the University’s Openness in Research Policy; (2) publication and broad sharing of research results; (3) faculty freedom to pursue research topics and methodologies of their choice; (4) conformance to the University’s primary mission of teaching and research; and (5) enrichment of the educational experience for students and postdocs.
The guidelines also draw a distinction between Affiliate Programs and sponsored research — the former does not involve a commitment to provide specific research reports, a budget or a specific work plan. Furthermore, Affiliate Program membership cannot involve special rights to intellectual property.
Another stipulation listed in Stanford’s establishment policies is that each Industrial Affiliate Program “create and maintain a public website that is operational and up-to-date.”
Despite this common set of rules, the framework of each affiliate program varies. For instance, some programs have hierarchical membership levels under which higher monetary contribution allows for access to more resources. At the highest level of membership, companies can receive exclusive benefits such as admission to governance boards, participation in mentorship programs and the ability to send visiting scholars to Stanford for longer periods of time.
The Natural Gas Initiative (NGI) Affiliates Program, for example, has three levels — Sustaining Member ($250,000 per year), Corporate Member ($75,000 per year) and Basic Member ($35,000 per year), with most participating companies falling under corporate membership. Top tier Sustaining Members are given a permanent seat on the NGI’s governance board, which “help[s] establish research priorities and recommend[s] projects for seed funding.” They can also institute a visiting scholar in the lab and become part of a program in which Ph.D. students and faculty visit their company and present data. Basic Members, on the other hand, are only given access to “informed research” and informal lab meetings.
Academia and industry: What’s the relationship?
Another concern emerging from Stanford’s AI Affiliates program is a potential conflict of interest private funding of research can create among the work of academic scholars.
John Ioannidis, co-director of the Stanford Meta-Research Innovation Center (which studies research) and a leading scholar on research policy, acknowledged the ever-more intertwined nature of industry and academia, and the potential risks that this phenomenon poses.
“Research can sometimes create conflicts of interests and distort research agendas, even subconsciously; for example, it’s been shown that studies sponsored by companies have a three times higher chance of reaching conclusions in favor of industry,” he said. The study he referenced, a 2010 paper by German doctor Wolfgang Becker-Brüser, demonstrated that several factors contributed to this phenomenon, from potential bias in study design to data manipulation.
However, he firmly believes that Stanford should not ban initiatives such as the Industrial Affiliates Program, and that generalizations that such programs subvert science are dangerous.
“Fruitful results can undoubtedly be reached at the intersection of industry and academia, as long as we ensure that there are clear and transparent rules surrounding funding,” Ioannidis explained.
As the lines between industry and academia weaken, industry leaders themselves are also discussing the potential risks of this phenomenon. At the recent talk “CRISPR, AI, and the Ethics of Scientific Discovery,” Stanford computer science professor and machine learning pioneer Fei-Fei Li warned that “the line between industry-sponsored and academic research has blurred a lot.” To ensure integrity and independence in academia, she argued that there should be a clear and consistent framework for governing any corporate-sponsored research.
Ioannidis, on the other hand, proposes that relationships be evaluated on a case-by-case basis. He believes that all affiliate program members should safeguard research independence and promote the sharing of knowledge. Rob Reich, a professor of political science and director of the Center for Ethics in Society, echoed this belief.
“The strings attached to any program should be transparent so that external observers can examine what is going on,” he said. “Teaching and research should not be dependent on funders’ interests.”
Contact Max Hampel at mhampel ‘at’ stanford.edu and Marianne Lu at mlu23 ‘at’ stanford.edu.
A quote from Rob Reich has been corrected to indicate his belief that teaching and research should not be dependent on funders’ interests, not fund interests. A previous version of this article also stated that Stanford’s E-Wear Initiative program does not have a functioning public website. In fact, the program does have a functioning public website (the article previously linked a non-functional site that was out-of-date). The Daily regrets these errors.