Last July 7th and 9th, Andrés Izquierdo participated with two interventions in the Second WIPO 2020 Artificial Intelligence (AI) Conversation. His topics where Deepfakes in AI, and Litigation in AI and Machine Learning.
The Second Session of the WIPO Conversation on IP and AI was held over three days from July 7 to 9, 2020 as a virtual meeting. ¨Over 2,000 people from 130 countries, including representatives of member states, academic, scientific and private organizations, followed the deliberations. More than 50 speakers from a diverse range of organizations spoke during the meeting, underlining the timeliness and magnitude of the discussion on IP and AI.¨
The Conversation was chaired by Ambassador François Rivasseau, Permanent Representative of France to the United Nations and other International Organizations in Geneva.
The First WIPO Conversation on AI and IP was convened by WIPO Director General Francis Gurry in September 2019 and ¨brought together member states and other stakeholders to discuss the impact of Al on IP policy, with a view to collectively formulating relevant questions.¨
Intervention on Deep Fakes
Watch video of intervention here
¨Thank you Mr. Francis Gurry, Dr. Ulrike Till, and Ambassador Rivarseau for the floor.
Today, I am here to talk about Deepfakes and AI and machine learning.
I want to provide some highlights from my research.
First, while Deep Fakes technologies can produce harmful effects on politics and society, it can also positively contribute to industries such as healthcare, agriculture, transportation, to name just a few.
Second, to combat the harmful effects of deepfakes, there are current legal tools that can be used, such as: defamation, fraud, intentional infliction of emotional distress, blackmail, publicity rights, and copyright.
Third, deepfakes legislation passed in California can serve to produce future case studies. Law AB 730 prevents improper influence on political elections, and law AB 602 forbids the unauthorized use of a person’s likeness in Pornography.
WIPO could study the implementation of these laws as a way to combat deepfakes without harming the further development of AI and machine learning.¨
Intervention on Litigation on AI and Machine Learning
Watch video of intervention here
This intervention was published by PIJIP, American University, Washington College of Law, see post here.
¨Today, I want to provide some highlights from my research on current litigation on AI and machine learning. Case law can provide guidance on how to address these new technologies.
I have reviewed litigation from the United States[1], Australia[2], China[3], and Canada[4] which involves AI and Machine Learning. Litigation regularly involves a diverse set of laws including copyright´s creativity, authorship, or registration requirements; publication and communication rights; violation of Technological Protection Measures (TPM’s); breach of contracts; trade secret misappropriation; or enforcement of website terms of service.
Ethical and human rights considerations have also been a matter of litigation. A Dutch Court[5] recently prevented the use of a technological system when it was judged unable to protect the right of privacy under the European Convention on Human Rights.
Where exceptions and limitations are the main focus, there is a series of cases in the US by which copying works for non-expressive purposes is justified as fair use[6]. Cases also show that copyright law would not apply for transient copies made during computational analysis[7].
The continued track of cases by the WIPO Artificial Intelligence and Intellectual Property Strategy Clearing House[8] will effectively help inform national policy reform.
Director General, colleagues, Thank you very much.¨
[1] UAB “Planner 5D” v. Facebook Inc., et al., No. 3:19-cv-03132, N.D. Calif. (involving copyright infringement and trade secret misappropriation claims over a home design website that allows users to create virtual interior design scenes using a library of virtual objects. The defendants proved their way out of court by the establishing the following: the plaintiff did not fulfill the court’s copyright registration requirements; the contented works did not have enough creative requirements to be copyrightable; the databases were not copyrightable either; the terms of service of the website that restricted any downloads were not enforceable to the defendant; and that there was no trade secret misappropriation).
[2] Acohs Pty Ltd v Ucorp Pty Ltd [2010] FCA 577 (involving subsistence of copyright in data sheets generated electronically. The plaintiffs alleged infringement of copyright, trade practices, and trademarks. The respondents denied copyright creativity requirements on the data sheets, and denied authorship of Acohs employees). See also IceTV Pty Ltd v Nine Network Australia Pty Ltd [2009] HCA 14 (involving copyright on computer generation of weekly TV program schedules).
[3] Shenzhen Tencent Computer System Co., Ltd. v. Shanghai Yingmou Technology Co., Ltd. (Involving A copyright infringement claim for an article written by the artificial intelligence software Dreamwriter. The discussion was centered on authorship requirements and plaintiff was awarded copyright and damages).
[4] Century 21 Canada Limited Partnership v Rogers Communications Inc, 2011 BCSC 1196 Century 21 (involving the indexation and linking to the Century 21 Website. Plaintiffs claims included breach of contract, trespass to chattels, enforcement of website's terms of use, and copyright infringement); Amel Chamandy / Galerie NuEdge Fine Arts v. Adam Basanta, 500-17-104564-185 (involving copyright and trademark infringement claims related to a project that uses artificial intelligence to automatically generate art images. Decision is pending on this case).
[5] NJCM c.s./De Staat Der Nerlanden (SyRI) before the District Court of the Hague, case no. C/09/550982/ HA ZA 18/388. (Holding by which the Dutch government was banned on SyRI, a technology used to detect possible social welfare fraud.
[6] Authors Guild, Inc. v. Google, Inc., 804 F.3d 202 (2d Cir. 2015), Authors Guild, Inc. v. HathiTrust, 755 F.3d 87 (2d Cir. 2014).
[7] Most recent case is Ticketmaster Corp. v. Tickets.com, Inc. (Holding that Ticket.com’s webcrawling software activity to
temporarily copy Ticketmaster's web pages to extract factual data about upcoming concerts and events was not copyright infringement). See also Cartoon Network LP v. CSC Holdings, Inc., 536 F.3d 121, 127 (2d Cir. 2008); Capitol Records, LLC v. ReDigi Inc., 910 F.3d 649, 657-58 (2d Cir. 2018); Soc'y of the Holy Transfiguration Monastery, Inc. v. Archbishop Gregory of Denver, Colo., 689 F.3d 29, 55 (1st Cir. 2012); Capitol Records, LLC v. Escape Media Grp., Inc., No. 12-CV-6646(AJN), 2015 WL 1402049, at *40 (S.D.N.Y. Mar. 25, 2015). For European Union see Art. 5(1) Info Soc Directive 2001 (establishing a mandatory exception for temporary acts of reproduction) and 5(3) Info Soc Directive 2001.
[8] See WIPO Artificial Intelligence and Intellectual Property Strategy Clearing House at https://www.wipo.int/about-ip/en/artificial_intelligence/policy.html#clearing_house
