The 2018 meeting of the International Technology Law Association (ITechLaw) was held March 7-9, 2018 in Hong Kong bringing together lawyers, regulators, business leaders, futurists and entrepreneurs to discuss the challenges they face today at the intersection of law and technology as well as offering their thoughts on what the future might hold.
Startups: Unique Businesses with Common Challenges
Hong Kong served as an appropriate background for this meeting as merely days prior to this event the government announced a HK$50B budget investment for driving technology startups in Hong Kong. Despite having a population of only 7.5 million people, it is home to over 2,200 startups and approximately 30 accelerators and incubators. The conference kicked off with presentations from six different startups based or operating in Hong Kong covering everything from artificially intelligent asset management to grocery delivery, to automated know-your-client compliance, to medical devices.
Three messages emerged from the discussions with the startups. First, across all sectors, data is the currency of the digital era. In every single presentation, companies were either monetizing data as part of their revenue stream or were actively considering whether at some point in the future it made sense for them to monetize the data they are collecting/generating. This prompted a lot of active questions, including (1) what data should companies collect, (2) whether pseudo-anonymous (“pseudonymous”) data adequately protects consumers both from a regulatory and ethical perspective, (3) how the soon-to-be enforced GDPR in Europe will impact this revenue stream, and (4) who really owns this data? However, none of these questions presented easy answers as the regulatory and business landscapes are still rapidly evolving and are not at all settled. Second, as artificial intelligence increasingly underlies the platforms and devices that consumer use, there is an increasing regulatory discomfort with the opacity of artificial intelligence. While it is relatively easy to ask a person to explain their thought processes for how they arrived at a decision, it is impossible (presently) to recreate the steps of a deep learning neural network to see how a conclusion is derived.
This has caused a lot of concern with regulators around the world, in both mature and developing markets, and has caused a lot of technology to be rolled out as an “augmentation” to a live professional rather than a platform that can interact with consumers directly. However, because regulations have not permitted the entire elimination of humans from the service delivery model, these businesses have also been somewhat hamstrung in their ability to deliver maximum efficiency. Third, improved data have allowed startups to micro-target populations that have been underserved or ignored historically. By combing these small populations across disparate geographies, startups are able to create a new market that is sufficient to sustain a business where previously there may not have been enough of a critical mass to sustain their company.
Keynote: Legal Technology Inside A Technology Company
Tim Steinert, General Counsel of Alibaba delivered the keynote address. His remarks focused on two areas, first on the way that Alibaba Group leverages technology in their business model (external facing technology) and second on the way that technology has impacted the way that Alibaba’s legal department works. Highlighting the way that the landscape has changed, even for a technology company, he noted that from its founding in 1999 through to 2010, Alibaba was only an e-commerce platform and had no presence on mobile devices. Since introducing its first mobile app (for TaoBao) in 2010, Alibaba has seen mobile grow to over 80% of its revenue generation. Looking toward into the future, he noted that Alibaba is fundamentally a data driven business and this is increasingly the most important asset the company possesses. This truth cuts across all of their business lines, from B2B and B2C ecommerce, to financial services in their Ant Financial arm. Looking to the future, Alibaba is also exploring even more product lines and business. First, taking note that software is increasingly being embedded into hardware, Alibaba has been exploring getting into chip design. Similarly, as the need for processing speed reaches nearly unthinkable levels (during the 11/11 “singles day” sales event in mainland China, Alibaba was processing 250,000 financial transactions per second), the company is exploring quantum computing solutions. Finally, Alibaba remains very interested in the “internet of things” and connected devices that continue to proliferate around the globe. Alibaba is considering entering smart-city, smart-manufacturing, and smart-retail business lines by leveraging identification, tracking, and embedded data gathering tools, supplemented with the data it already has as well as additional data it may be able to collect.
Turing to Alibaba’s legal department, Steinert noted that he has created a significant legal operation function with 10 full time employees and an additional 60-70 developers working on technology to help his legal function operate more efficiently and effectively. “Too many smart people doing too many stupid things” summed up his view on the amount time his legal department spends answering repeated questions, chasing down signatures, or examining contracts to see whether they contained a “right-of-first-refusal” clause in them or not. In the time that Steinert has been at Alibaba, its legal department has grown from 35 to 450 individuals and now operates and sits around the globe. His primary areas of focus for improvement in his legal department are (1) company secretarial since Alibaba has over 1300 legal entities to manage around the world (2) responding to KYC requests, including digital signature and verification (3) litigation management for high volume (~15,000) but low value (<1,000RMB) matters (4) content management for non-monetary aspects of executed agreements (5) external counsel management (6) tracking legislative affairs and government affairs work on influencing regulations (7) monitoring Alibaba’s minority investments in over 200 other companies. Despite these challenges, he also noted that chat bots had been successfully implemented in several areas to offer internal clients a self-service portal for routine questions and has liberated his lawyers from dealing with these requests.
After the impact and use of data, the second major theme of the conference was artificial intelligence and is impacts on law and business. The expert panel on AI first dispelled a common fear regarding AI – that AI would somehow turn malevolent and bring about a cyber apocalypse (this fear being spurred by recent comments from Elon Musk). The key thing to keep in mind is that there is no such thing as generalized AI at this time. Where AI exists, it is targeted to a very narrow and very specific task or set of tasks. There is no such thing as a machine that has sentience and there is no such thing as AI that has the ability to self-direct its own goals or select the task that it wishes to accomplish. That does not mean that there are no problems with AI. For example, using AI in military applications poses many thorny ethical and moral issues. But for now, the robot uprising remains solidly within the realm of speculative fiction.
Turning to the more immediate impact of AI on the practice of law, the products that have been introduced thus far have been focused on areas such as legal research, contract/document generation, due diligence and e-discovery. It was also noted, that a form of invisible AI in legal technology is in how the user interface for many big companies such as Thompson Reuters now tracks user actions and uses this monitoring to improve the way results are displayed. One key takeaway regarding AI, whether in the healthcare space, the legal space, or any other application, is that the sophistication of the insights that AI can provide is directly correlated to the amount of data the system processes. For high volume data sets, AI is able to deliver the greatest returns and hold the potential for uncovering the most hidden insights.
Building on the above point regarding the large appetite for data that AI systems bring with them, panelists considered whether this posed a hidden competition law risk. Large companies, whether Google or Thomson Reuters, have the advantage of sitting on massive data sets to being with. This can be extremely difficult for a start-up to compete with and thus can limit the number of entries into this space. The panel experts noted that thus far, startups are proliferating in all segments of AI development and there are no signs that this trend is slowing down. Additionally, the amount data that is needed to train an effective AI system has been on a steady and continuous decline for the last decade. As a result, if a startup is able to obtain a high-quality, albeit small, data set they are able to create highly effective tools. If this trend continues, any advantage presently entertained by companies with large proprietary data sets will evaporate.
Another area of concern for AI is in areas where it has not merely augmented human decision making but has even started to supplant human decision making. The example discussed was the COMPASS software used in the US to determine likely recidivism rates as an aid to judges in criminal sentencing. COMPASS has come under fire for having embedded biases, learned from historical data sets used to train the system which then become self-reinforcing. While this is undoubtedly a problem, there are two things to keep in mind. First, this bias was detected and demonstrated, which means that we have the ability to police the machines we create and continue to improve and refine them. Second, it is inappropriate to compare AI to a standard of perfection. In most cases, humans are actually very poor at decision making and can be extremely biased, without possessing any self-awareness of our own biases (relying on the work of Daniel Kahneman). The goal of people working in AI is not to create a perfect system. It is merely to create something that is better than the human equivalent.
Looking to future developments in AI, perhaps the most exciting area of academic research is in Explanatory AI or “Ex-AI”. This is the combination of two neural networks to work in tandem. This emerging technology attempts to solve the “black box” problem of AI. One neural network serves as the decision-making function, and the other neural network helps to explain the reasoning and decision calculus behind how the decision was made by the other system. This technology is still in its infancy but as it matures it may hold the potential to solve some of the regulatory concerns that underlie direct use of AI tools by consumers, who may not be very sophisticated.
When it comes to the legal sector and AI, two problems remain. First, there is a massive disconnect between the interest in AI and the willingness to adopt AI. While some legal departments and law firms have made the move, the overall sector has chosen to mostly stand on the sidelines and observe how the technology develops. Second, AI represents a change in the way legal work is performed. In other words, a process change. And whenever there is a process change, there must be a culture change. Law firms culture is notoriously static and resistant to change. However, it is not impervious to change. As both the volumes of data increase and the technology matures, it seems that AI’s infiltration of the global legal practice is all but inevitable.
Global Innovation, Regional Issues
The final major theme of the conference was the challenges that technology companies and lawyers face as they increasingly attempt to traverse borders and deal with sets of law that are still firmly rooted in a soil-based understanding of jurisdiction and a physical approach to verification and trust.
As an example, across Asia different countries have seen various levels of adoption of e-payment systems. Some of this variation may be historical. For example, in Hong Kong, credit cards are much more accepted and ubiquitous than they are in mainland China, and thus they are harder to displace.
However, there are also issues related to the legal framework. For example, if a small restaurant were to adopt an e-payment from a bank or other service provider, the corporate chairperson plus two additional signatories would be required to enter into the payment agreement with the e-payments vendor. And these signatures must be physical and accompanied by the company seal (chop). In order for an e-vendor to sign up a single small restaurant requires multiple meetings and convening a small meeting involving at least four people. Then, to do this on any level of scale becomes a massive undertaking due to this burdensome regulation.
Another example came from India. Although the Indian federal government has passed e-signature and e-KYC laws that have paved the way for a simplified electronic platform for personal loans, local states in India still require physical paper loans to be stamped as part of the stamp tax collection process. This effectively wipes out the efficiency and paperless benefits that an e-loan system delivers in the first instance.
On a global level, there is quite a bit of speculation regarding the GDPR in Europe. Perhaps the most common speculation is that the GDPR will become the de facto defining global standard. As the thinking goes, since the GDPR is so limiting, companies will choose to comply with the GDPR and assume that by complying with the strictest law they are automatically complying with the laws of all other jurisdictions. Many lawyers in attendance were quick to point out the pitfalls in this approach, particularly because although the GDPR may be the “strictest” global regulation, there are still specific types uses of data that are permissible under the GDPR yet not permitted in all jurisdictions around the globe. This regulatory challenge presents a lot of work for lawyers, and a major challenge for many companies wishing to enter additional jurisdictions.
The world’s most populous country, China, also poses a unique set of challenges. One key to any technology company is its proprietary intellectual property. China has come a long way in recognizing and protecting intellectual property, even going so far as to set up specialized IP courts. Interestingly, these specialized courts were favorably reviewed by lawyers that had appeared before them for both their expertise and impartiality. China still poses challenges, not the least of which being that its cloud network is not secure (by international standards) and China still imposes many restrictions on licensing and non-Chinese ownership/investment. China’s purposely vague states secrets law also poses a challenge to businesses who may find themselves falling within the scope of the law despite being solely involved in purely commercial activities directly with consumers. Because of the Great Firewall, the average Chinese consumer is unable to access Facebook, Twitter, Google, YouTube, and many other Western technology products. However, China is well aware that the wealthy in China can, and regularly do, access content that would otherwise be blocked by the Firewall. Accessing this content is permitted at most times, although it will be shut down during politically sensitive times such as during the “Two Sessions” meetings in Beijing. For now, the technology environment in China is both fast-paced and growing due to its prestige, public and private investment, and academic support.
However, the technology environment in China also remains severely ring-fenced by Beijing and very risky for Western companies to enter.
Conclusion: Different Perspectives, Common Goals
As is so often the case, the world’s regulatory systems lag behind the pace of technology in everything from the development of cryptocurrencies to AI to e-signatures. Nevertheless, technologists, futurists and lawyers are boldly charging ahead in developing the next platforms, tools and devices that will change everything from the way we eat, relax and stay healthy to conduct business, comply with contracts and even practice law. The optimism and resolve throughout the conference was unmistakable.
While there are many challenges, there are even more talented, thoughtful and dedicated people working to resolve these challenges.
For more see Conventus Law.