Security, Privacy and Innovation – Reshaping Law for the AI Era

Earlier this month and final month, the National Security Commission on Artificial Intelligence, the Reiss Center on Law and Security at NYU School of Law, the Berkman Klein Center for Internet & Society at Harvard University, and Just Security convened a three-part digital symposium of specialists to debate crucial authorized points round the rising use and affect of synthetic intelligence (AI). Titled “Security, Privacy and Innovation: Reshaping Law for the AI Era,” the symposium comprised three periods convening main students, practitioners, and thought leaders on a few of the most tough and pressing aspects of the AI period. In case you missed the occasion, this recap describes highlights from every panel. Further particulars of the symposium will be accessed right here. These descriptions and observations are our personal and aren’t essentially shared by every of the panelists.
September 17 – Responding to AI-Enabled Surveillance and Digital Authoritarianism

Jonathan Zittrain, Faculty Director of the Berkman Klein Center for Internet & Society, moderated the first panel. The panel featured Olufunmilayo Arewa, Murray H. Shusterman Professor of Transactional and Business Law at Temple University’s Beasley School of Law; Chinmayi Arun, Resident Fellow at Yale Law School; Ronald Deibert, Director of the Citizen Lab at University of Toronto’s Munk School of Global Affairs & Public Policy; and Ambassador Eileen Donahoe, Executive Director of Stanford University’s Global Digital Policy Incubator.
The panelists targeted on a number of, interconnected areas of concern, together with the fast growth of AI applied sciences paired with the inapt authorized safeguards of the previous, the use of AI applied sciences to perpetuate human rights abuses, and the world nature of those points. As Deibert put it, “all of us stay on this new type of world ether of knowledge that’s related to however separate from us.” Each panelist drew out distinctive angles of the potential harms of AI. Arewa highlighted the potential for abuse resulting from focus of energy in know-how firms like Facebook and Google. Arun spoke about how datafication of individuals can result in erasure of sure teams, giving the instance of how datafying individuals as male or feminine erases those that don’t determine alongside the gender binary. She additionally spoke about how, when coping with cross border questions, worldwide legislation “affords us highly effective norm setting,” however largely doesn’t create accountability for highly effective know-how firms.
Ambassador Donahoe narrowed in on AI applied sciences deployed by authoritarian regimes to “form citizen motivation and conduct,” like China’s social credit score system. She famous that such applied sciences “not solely violate privateness and civil liberties, however they actually undermine human company and go to the coronary heart of human dignity.” Her greatest concern was the menace of digital authoritarianism as a governance mannequin, particularly because it spreads throughout the world and competes with democracy.
The panelists additionally offered steering on how civil society and governments in democratic states can deal with the dangerous results of AI on a number of ranges. In the worldwide enviornment, Ambassador Donahoe argued that “on the democratic aspect, we’ve got mainly failed to offer a compelling various” to the digital authoritarian regime. She laid out a three-part geopolitical framework, which she subsequently elaborated on right here. The three parts had been: develop a democratic governance mannequin for digital society, spend money on values-based worldwide management, and win the technological innovation battle to maintain energy in democratic states. Deibert advised constructing momentum in countering abusive “despotism as a service” practices by enhancing home oversight of surveillance firms and applied sciences, beginning with businesses equivalent to the U.S. National Security Agency (NSA) and Canada’s Communications Security Establishment (CSE).
Arewa articulated a framework for regulating personal actors primarily based on each transparency and legal responsibility, however she additionally acknowledged the impediment of regulatory seize, even in international locations that revered the rule of legislation. On the transparency aspect, she gave the instance of how Apple’s app monitoring transparency led to many fewer customers opting to be tracked. On the legal responsibility aspect, she identified that though Mark Zuckerberg depends on the legal responsibility limitations embedded in company legislation, that might not be acceptable for somebody like him who serves as an organization’s controlling shareholder and CEO and additionally sits on the board.
Arun favored an strategy that adopted laptop scientists’ analysis to know how “accountability will be hardwired into the constructing of those methods.” In addition to anticipating harms, she advocated for monitoring every use of AI and creating mechanisms to stroll again any dangerous results.
The panelists concluded by articulating hopes for a future the place democratic values are infused into AI know-how.
September 24 – Constitutional Values and the Rule of Law in the AI Era: Confronting a Changing Threat Landscape

Julie Owono, the Executive Director of Internet Sans Frontières (Internet Without Borders), a member of the Facebook Oversight Board who can also be affiliated with the Berkman Klein Center on Internet & Society, moderated the second panel. The panel featured Glenn Gerstell, a senior advisor on worldwide safety with the Center for Strategic & International Studies and the former common counsel for NSA; Aziz Z. Huq, the Frank and Bernice J. Greenberg Professor of Law at the University of Chicago Law School; and Riana Pfefferkorn, a Research Scholar at the Stanford Internet Observatory.
The dialog targeted broadly on how the American constitutional system is challenged by many emergent issues with the use, growth, and deployment of AI. As a place to begin, Gerstell described AI instruments as “crucial, pervasive, and problematic.”
The panelists mentioned how overseas adversaries like China have invested closely in AI applied sciences to intently surveil their very own populaces, achieve a aggressive edge in the world market, and rapidly kind by way of intelligence. In order to maintain up with technological innovation and defend necessary nationwide safety pursuits, the United States should proceed to develop and depend on AI. But the panelists emphasised that present authorized parameters don’t sufficiently defend the privateness pursuits of on a regular basis Americans or present sufficient protections and cures for dangerous actions by governments or personal firms.
As for the present authorized construction, the panelists targeted on the restricted protections supplied by the Fourth and Fourteenth Amendments. They agreed that the Fourth Amendment offers a restricted guardrail round the use of AI applied sciences by nationwide safety establishments and acknowledged that necessary questions nonetheless exist about whether or not AI can provide rise to possible trigger for a warrant. Gerstell identified that the most related case about the limits of presidency surveillance underneath the Fourth Amendment, Carpenter v. United States, offers little on-point steering about the limits on knowledge the authorities can accumulate.
Huq asserted that the Fourteenth Amendment was unable to deal with the most urgent considerations about authorities use of AI, equivalent to disproportionately excessive false positives that negatively affect racial minorities and girls. While the Equal Protection Clause prohibits governmental actions primarily based on racially discriminatory intent, AI applied sciences, he argued, are not often designed with the intent to discriminate; as a substitute they incorporate biases by way of negligence or inattention.
Pfefferkorn additional defined the quite a few challenges posed by AI in a felony justice context, the place prosecutors could also be unable to totally clarify AI applied sciences used to gather or analyze proof towards defendants. This could also be as a result of the know-how is opaque even to its inventors, or as a result of contractual or nationwide safety obligations stop the distributors from disclosing how the instruments function.
The panelists additional identified that the menace posed by the use of AI comes not solely from the authorities however from firms that aren’t certain by constitutional limitations. The energy and worth of AI applied sciences require gathering huge quantities of knowledge about people, and this knowledge typically comes from these firms’ customers. Accordingly, the panelists contended {that a} rights-focused framework is insufficient in the context of the threats posed by AI.
The panelists pressured the pressing want for laws that extra clearly delineates privateness rights for Americans, defines who can accumulate their knowledge in public areas and what that knowledge can be utilized for, and bans some AI functions in notably delicate areas. Pfefferkorn identified that privateness laws and doctrine from the Sixties and Seventies are far behind the present technological capabilities in the present day, and that altering know-how could correspond to a change in the definition of affordable expectation of privateness. Huq advocated for a federal company much like the FDA or CDC with administrative authority to manage the AI trade; nevertheless, he cautioned that political will for such an company doesn’t exist.
October 1 – Protecting and Promoting AI Innovation: Patent Eligibility Reform as an Imperative for National Security and Innovation (Panel 1)

Ruth Okediji, Jeremiah Smith, Jr. Professor of Law at Harvard Law School, moderated the first panel, which featured Paul Michel, former Chief Judge of the Federal Circuit; Andrei Iancu, former Undersecretary of Commerce for Intellectual Property and former Director of the U.S. Patent and Trademark Office (USPTO); and David Jones, Executive Director of the High Tech Inventors Alliance.
Okediji launched the subject of patent eligibility reform, and famous the National Security Commission on Artificial Intelligence’s last report was launched in March 2021. That report features a non-exhaustive record of 10 mental property-related issues for the United States to evaluate as a part of its nationwide safety technique. One of these issues is patent eligibility reform.
Judge Michel offered crucial background on the difficulty. Patent eligibility is considered one of the threshold necessities for a patent to be granted — or for an issued patent to be upheld when challenged in litigation. Under Section 101 of the Patent Act, 4 broad classes of innovations are patent-eligible: processes, machines, manufactures, and compositions of matter. According to Judge Michel, the Supreme Court’s choices in Mayo v. Prometheus (2012) and Alice Corp. v. CLS Bank International (2014) modified the patent eligibility panorama. These choices expanded the scope of three judicial exceptions — legal guidelines of nature, merchandise or phenomena of nature, and summary concepts — to the 4 statutory patent-eligible classes talked about above. Judge Michel opined that, previous to 2012, the U.S. patent eligibility regime was clear and constant, and challenges to eligibility had been uncommon, however eligibility challenges have develop into commonplace since Mayo and Alice. Meanwhile, 27 European international locations and many Asian international locations have considerably broadened their patent eligibility standards, and a whole lot of patents deemed ineligible in the United States have been deemed eligible elsewhere. Judge Michel concluded his remarks by calling for congressional reform of the U.S. patent eligibility regime. Reform efforts in 2019 stalled, however some discussions on Capitol Hill are presently underway once more.
Iancu and Jones then engaged in a spirited debate. Iancu usually agreed with Judge Michel that the legislation of patent eligibility is in a state of unpredictability. He argued that the personal sector would require better readability and certainty from the patent system in an effort to really feel incentivized to innovate and spend money on new, disruptive applied sciences, equivalent to AI. AI-related innovations have typically been rejected by the present patent regime, which continuously views them as mathematical formulation and “summary concepts.” According to Iancu, the process for defining “summary concepts” and figuring out whether or not a selected invention ought to be patent-eligible or not remains to be unclear. He highlighted new tips that the USPTO issued in 2019 to synthesize court docket choices and present an analytical framework for patent eligibility analysis. But this alone isn’t adequate, he acknowledged, calling on Congress to reform the eligibility statute itself, which was written in 1790, when applied sciences equivalent to blockchain, AI, and quantum computing couldn’t have been fathomed.
Jones, on the different hand, argued that the present regime is working nicely and spurring innovation. He cited, for instance, an empirical research that demonstrated that firms elevated their analysis and growth (R&D) investments after Alice as a result of they may not merely depend on patents for applied sciences that had been now not eligible for safety. Limiting the scope of eligibility is useful, he advised; patent candidates shouldn’t be capable of merely add “the magic phrases, ‘on a pc’” and declare an summary concept to be patentable. Jones additionally argued that the post-Mayo/Alice regime has been pretty predictable, and that patent candidates have tailored in a short time to modifications in the jurisprudence.
The panelists additionally mentioned U.S. patent eligibility particularly in the context of nationwide competitiveness. Jones defined that underneath the TRIPS settlement — which has been described by the WTO as the “most complete multilateral settlement on mental property” — the signatory international locations (which comprise most of the world) are obligated to deal with overseas inventors and home inventors in the identical method. Thus, firms won’t essentially migrate their R&D efforts away from the United States (if they’re in search of U.S. patents), Jones argued. On the different hand, Judge Michel warned that “capital is fleeing the United States and fleeing laborious know-how for much less dangerous investments.” Iancu mentioned that the United States must do extra to incentivize startups, small-and-medium enterprises, and enterprise capital corporations to spend money on disruptive applied sciences right here at dwelling in an effort to match competitors and innovation from China. He argued that offering sufficient protections for patents is a approach of making these incentives. Jones countered that some research present there was a rise, not a lower, in startups’ entry to enterprise capital funding in the aftermath of Alice.
Okediji concluded the first panel by noting the significance of discussing these points for nationwide competitiveness and issues of what will be executed with patent levers.
October 1 – Protecting and Promoting AI Innovation: Patent Eligibility Reform as an Imperative for National Security and Innovation (Panel 2)

Kristen Jakobsen Osenga, Austin E. Owen Research Scholar & Professor of Law at the University of Richmond School of Law, moderated the second panel. The panel featured Ryan Abbott, Professor of Law and Health Sciences, University of Surrey School of Law, and Adjunct Assistant Professor of Medicine at UCLA’s David Geffen School of Medicine; Drew Hirshfield, who’s presently performing the capabilities and duties of the Undersecretary of Commerce for Intellectual Property and Director of the USPTO; Hans Sauer, Deputy General Counsel and Vice President for IP at the Biotechnology Innovation Organization; and Laura Sheridan, Senior Patent Counsel and Head of Patent Policy at Google.
Osenga opened the dialogue by noting that the panel would develop on the first panel and talk about some on-the-ground, sensible implications of patent eligibility points.
In their opening remarks, every panelist shared their preliminary observations on patent eligibility points. Hirshfield known as for better predictability in patent eligibility legal guidelines, a extra environment friendly course of for evaluating patents, and a nationwide technique for defending AI. Sheridan opined that the present patent eligibility regime is balanced and supportive of AI innovation. “Any disruption of the steadiness would really hurt innovation and rising applied sciences, not assist it … patenting in AI is definitely flourishing, regardless of what the [National Security Commission on AI] report says,” she argued. Sauer famous that international locations round the world pay shut consideration to U.S. patent legislation, together with any systematic divergences in outcomes in the United States versus elsewhere. “We have lived with a disparate state of affairs,” he mentioned, referring to the biotech trade’s challenges in acquiring patents in the United States in comparison with different international locations. Abbott spoke to the variations between AI’s disruptiveness and earlier generations of applied sciences, notably emphasizing AI’s distinctive means to generate its personal artwork, music, and innovations. How the U.S. patent system treats AI-generated innovations (in comparison with conventional, human-invented IP), Abbot noticed, can have necessary authorized and financial ramifications in the years to come back.
Commenting on the present panorama of patent functions, Hirshfield famous that 18 to 19 % of functions to the USPTO now have some type of AI in them. Recognizing the tendencies, the USPTO is enterprise a spread of initiatives associated to AI, he mentioned. He additionally spoke of the challenges emanating from an absence of readability in patent eligibility jurisprudence — and raised considerations about what which may imply for the AI improvements of tomorrow.
Sheridan added that Google has inspired the USPTO to offer a sturdy technical coaching to its patent examiners, such that examiners can keep up-to-date on rising applied sciences. She additionally talked about that Google’s choice on whether or not to maintain an invention a “commerce secret” shouldn’t be primarily based on patent eligibility legislation; fairly, it’s primarily based on enterprise and product-driven issues, the nature of the know-how, and whether or not Google is snug with disclosure.
Sauer advised that U.S. patent legislation, because it stands, might probably invite copyists given the lack of clear protections, and that sure biotech patents could be higher protected in China. He additionally famous that the increased bar for patentability in the United States is main the biotech trade — notably diagnostics firms — to focus its investments extra on applied sciences that may be stored confidential (i.e., commerce secrets and techniques) or on instruments utilized in the R&D course of.
Abbott argued that the legislation, even because it stands in the present day, ought to permit for patents to be awarded for AI-generated innovations. He acknowledged that there’s presently a break up on this query in jurisdictions throughout the globe. “[T]he Patent Act was designed to encourage technological progress and producing socially priceless actions, and that that is precisely the form of exercise that patent legislation was meant to accommodate, and studying the legislation with that objective in thoughts, there isn’t any principled motive that an AI couldn’t invent one thing and that somebody couldn’t get a patent on that form of factor,” he mentioned.
In concluding, Sauer mused that we could sometime witness “a battle of AIs,” with AI-generated IP being scrutinized by an AI-driven patent company analysis course of.

Recommended For You