Intel­lec­tu­al Prop­er­ty in a Cyber World

Intel­lec­tu­al prop­er­ty is the legal asset in inno­va­tion. The val­ue of inno­va­tion to a busi­ness depends on the extent to which that inno­va­tion is direct­ed to the needs of the rel­e­vant mar­ket. If an inno­va­tion meets that need effec­tive­ly and unique­ly, it can pro­vide that busi­ness with sub­stan­tial­ly increased rev­enue and mar­gin. Inno­va­tion can there­by be the key source of busi­ness growth and prof­itabil­i­ty. The intel­lec­tu­al prop­er­ty in that inno­va­tion there­fore needs to be iden­ti­fied and protected.

The crit­i­cal nature of intel­lec­tu­al prop­er­ty is just as rel­e­vant in a cyber world as it is in a phys­i­cal world. How­ev­er, the tech­nol­o­gy inher­ent in a cyber envi­ron­ment can chal­lenge the fun­da­men­tal con­cepts of intel­lec­tu­al prop­er­ty law. Infor­ma­tion tech­nol­o­gy (IT) and arti­fi­cial intel­li­gence (AI) raise par­tic­u­lar issues for copy­right and patent law.

Copy­right sub­sists in a unique work expressed in mate­r­i­al form. It is an accept­ed prin­ci­ple of copy­right law that the elec­tron­ic rep­re­sen­ta­tion of a copy­right work con­sti­tutes an expres­sion in mate­r­i­al form. How­ev­er, under sec­tion 32 of the Aus­tralian Copy­right Act (Cth) (1968), copy­right can only sub­sist in a work cre­at­ed by an ​“author” which implies the input of intel­lec­tu­al effort. This rais­es the ques­tion as to whether copy­right can sub­sist in works cre­at­ed by the data process­es of com­put­er algo­rithms. AI has been used to cre­ate draw­ings, tables and even music. In one instance, an entire musi­cal pro­duc­tion has been devised based entire­ly on AI. More impor­tant­ly, AI can have the capac­i­ty for a ​“deep learn­ing”, func­tion under which arti­fi­cial neur­al net­works have been devel­oped to mim­ic the net­works of the human brain in order to process data.

The expo­nen­tial growth of AI can pose fun­da­men­tal issues for the tra­di­tion­al con­cept of copy­right. In Aus­tralia, the gen­er­al posi­tion is that ​“intel­lec­tu­al effort” must be involved in the cre­ation of a work in order for copy­right to sub­sist in that work. In Ice TV Pty Lim­it­ed v Nine Net­work Pty Lim­it­ed (2009) 239 CLR 458, it was held that the cre­ation of a tele­vi­sion pro­gram timetable via an algo­rithm con­sti­tut­ed use of the rel­e­vant algo­rithm ​“as a tool”. The author of the work was there­fore the per­son using that tool. In Acohs Pty Lim­it­ed v Ucorp Pty Lim­it­ed (2012) 201 FCR 123, the Full Fed­er­al Court found that copy­right did not sub­sist in data sheets cre­at­ed by a com­put­er pro­gram because there was no human inter­ven­tion in that creation.

The Aus­tralian posi­tion on the neces­si­ty for human inter­ven­tion in the cre­ation of a copy­right work is broad­ly con­sis­tent with US and EU law. How­ev­er, in the UK there is a spe­cif­ic pro­vi­sion in Sec­tion 9(3) of the Copy­right Design and Patent Act (1988) (UK) which states that: ​“in works which are com­put­er-gen­er­at­ed, the author will be tak­en to be the per­son by whom arrange­ments for the cre­ation of the work are under­tak­en”. Sim­i­lar leg­isla­tive pro­vi­sions exist in New Zealand, Hong Kong, India and Ire­land. A com­pa­ra­ble leg­isla­tive pro­vi­sion has been rec­om­mend­ed in Aus­tralia by the Copy­right Law Review Com­mit­tee but it has not yet been enact­ed in leg­is­la­tion in Australia.

The fore­go­ing UK pro­vi­sion inher­ent­ly recog­nis­es the sub­sis­tence of copy­right in machine-gen­er­at­ed works but as the ​“author” becomes fur­ther sep­a­rat­ed from the cre­ative process, it becomes more dif­fi­cult to iden­ti­fy the ​“intel­lec­tu­al effort” involved. This dif­fi­cul­ty is mag­ni­fied with regard to works cre­at­ed by ​“deep learn­ing” machines that effec­tive­ly pro­duce cre­ative works inde­pen­dent­ly of any human involve­ment. It has even been argued that the true author of a work cre­at­ed by a ​“deep learn­ing” machine is the machine itself. Such a notion strikes at the very heart of the tra­di­tion­al con­cept of copy­right and rais­es fun­da­men­tal issues as to the mean­ing of the ​“author” of a machine-gen­er­at­ed work.

Inven­tions cre­at­ed by AI also pose fun­da­men­tal issues for patent law. AI is cur­rent­ly being used by the World Intel­lec­tu­al Prop­er­ty Organ­i­sa­tion (WIPO) to assess patent claims. AI offers real advan­tages to WIPO Exam­in­ers in their analy­sis of the pri­or art to assess the nov­el­ty of an inven­tion. AI can even be used to make assess­ments of the patentabil­i­ty of an inven­tion in the light of decid­ed case law. In this respect, AI is enter­ing the realm of pro­fes­sion­al legal advice, tra­di­tion­al­ly pro­vid­ed by lawyers.

AI can also be employed to pro­vide options and vari­a­tions of exist­ing inven­tions to osten­si­bly cre­ate new and patentable inven­tions. How­ev­er, the devel­op­ment of such deriv­a­tive inven­tions by AI rais­es the issue of the ​“obvi­ous­ness” of such inven­tions. The patentabil­i­ty of an inven­tion is gen­er­al­ly assessed by Exam­in­ers accord­ing to the stan­dard of ​“per­sons skilled in the art” and it can be assumed that such Exam­in­ers have access to AI tech­nol­o­gy in mak­ing their assess­ments. If an inven­tion can rea­son­ably be derived by use of AI tech­nol­o­gy, an Exam­in­er, equipped with sim­i­lar AI tech­nol­o­gy, might there­fore find that the inven­tion is not patentable because it was ​“obvi­ous”. In this regard, Ai can be a two-edged sword for prospec­tive patentees.

The pro­lif­er­a­tion of IT in mod­ern soci­ety is occur­ring at an expo­nen­tial rate. The con­cept of the Inter­net of Things, involv­ing Human to Machine and Machine to Machine inter­ac­tion, is rapid­ly lead­ing to the cre­ation of Big Data under which the empha­sis is shift­ing from the own­er­ship of the tech­nol­o­gy to the own­er­ship of the data itself.

At the same time, there has been a pro­found shift in mod­ern soci­ety from an empha­sis on phys­i­cal cap­i­tal to an empha­sis on intel­lec­tu­al cap­i­tal. A few decades ago, the largest cor­po­ra­tions in the Unit­ed States were vehi­cle man­u­fac­tur­ers, mining/​resources com­pa­nies and finan­cial ser­vices sup­pli­ers. Today, the five largest com­pa­nies by cap­i­tal­i­sa­tion in the Unit­ed States are Apple, Alpha­bet, Microsoft, Ama­zon and Face­book. The infor­ma­tion tech­nol­o­gy of the fore­go­ing com­pa­nies per­vades vir­tu­al­ly all sec­tors of economies world­wide. Iden­ti­fy­ing and pro­tect­ing the intel­lec­tu­al prop­er­ty in this infor­ma­tion tech­nol­o­gy and data presents a com­pelling chal­lenge for lawyers.