Good afternoon everyone Thanks for coming To This panel which is titled or Human Rights Data governance and sandboxes for Ai in latin America and the Caribbean the panel is the result of the merge of two proposals One of them by derechos digitales and another by Data spere initiative And I hope that during the discussion We Will be able with this brilliant Group of Invitees to Navigate this hopefully false dichotomy between innovation and Human Rights Uh My name is jamila venturini I'm co executive director at dichos digitales which is a latin American civil society organization that
works at the intersection precisely of Human Rights and Technology Our work includes research advocacy public campaigning and also capacity building on these topics and We have been dedicated In The Last Four Years at least to understand How Ai is Being implemented and regulated in latin America we decided to dedicate efforts to that as we identified that the International discussions around Ai didn't include much evidence or much information and contextual information about What happened in latin America so We have been developing different typ types of studies to contribute with more evidence on Our context to try
to Reach Uh Standards that are useful for Our to We have developed Case studies and we Are about to launch um A document On regulatory sandboxes Which are key part of Our discussion today and in the end of the conversation I can share the Website Where you can find All the this this information so Going To Our topic today sandboxes are gaining ground in latin American countries as a regulatory tool for artificial intelligence With discussions for experimentation and normative advances ongoing in different Uh countries such as brazil Chile Colombia and Mexico While there are Still
questions about How to deploy them um They are they continue to Advance in the region as I said We have an Amazing Group of panelists today to present their perspectives on the topic and to help Us Navigate and iden lessons learned And ideas to guide Us and inspired Our countries as they Advance toward exploring these type of mechanisms We Will have Two rounds of interventions Uh followed by your questions so please Take Note of your question so we can get back to them during the conversation so I would like to Start with Felipe Uh Felipe
Mora is the founder and ceo of Vinci ventures Felipe holds a Masters Dee from the University of California a bachelor degree in International Relations from the University of brazilia and currently pursuing an executive mba at fundation Don cabr Felipe is a member of the brazilian institute of governance and he also serves as Vice President of the brazilian Ai association and President of fdc Angels please Felipe the flo iss Thank you and It's a great honor to be here um We have some guests online as well so Hi to All Of You um It's Uh It's
two very interesting topics And I like The way that you framed it um It's Uh an apparent um contradiction but It's not right and That's something that we need to Explore um something that I Always like To talk about in my um when I'm First starting um especially on on on audiences that Are Not so accustomed to the whole discussion on Ai is is a a really good study from um mit And I don't know How many Of You are familiar with it raise your hands if you are um It's called the Moro Machine have
you Heard about it you have um So The Moro Machine is is started I think in 2014 I'm not sure so It's probably at least 10 years old right now they have formally published their results but um If You're curious you can go on their Website It's translated to a bunch different languages including portuguese and you can do the test yourself so It's it runs different scenarios of the trolley problem so the trol problem Where you know if you don't do anything Uh Five people Die but if you Pull A lever then One person dies
but you get to feel responsible for that One person that Died Because you pulled the lever right Or You Let Five people Die so either way You're responsible um and so the moral Machine um study runs 40 I think 40 or or more different scenarios variations of that so What if you have to choose between killing the passengers versus the pedestrians What if you can choose between killing an elderly Woman Or a young Child What if It's a human versus a Cat and so on so It's different scenarios I don't recommend doing it on on
On A full stomach um It's It's It's a very heavy kind of exercise um and what I do like about that and thinking about the results is that it shows clusters of of Data look preferences so if you Look at the Western Side versus the eastern side of the globe Uh there seems to be some clusters that that that would Uh and This is speaking in very broad terms Um you would choose to save the elderly in one part of the world you choose to save the younger ones in another Part the humans versus the
Animals and so on and so even as we try to think About what are and we Had an interesting discussion in one of the panels this morning even If We try to think About what are the global things that we can think about and we can all agree on on very basic literal life and Death situations As a globe we can't find common ground but we can find common ground on the in the sense That there are priorities per region and there are things that we need to think about and Those are tough decisions right
um and so we as societies we need to Explore um as as collectives What are the things that we Uh want in scenario a or scenario B and some companies have actually taking a Step And I I'm Not Going To say the name of the Company Because I I Always Confuse the two but there's there's a German Auto manufacturer that said Uh if you buy the autonomous car from me you buy it knowing that We Will Always choose to save the passengers the Driver and the passengers so you you if you bu their System you
acknowledge that it's on the terms of service right um and we don't know What's the right answer but I think the Beauty of this of the sandboxes as a tool is that we get to find out we get to experiment of course Hopefully not in Life or Death situations right um but we get and and um understanding What is public interest understanding what We want as a society as a collective um and What We are technically able to do capable of doing and how we want to to possibly regulate What are Our values we can
test hypotheses um and Rather than than simply saying you know This is what europeans are doing And I want to import that from my country or This is what the Americans are doing or This is what the chinese are doing no no no Let Us Sit down right and have a and have and have a thought and experiment What is it that We actually want and the there's both a Beauty and also a a a um It's not risk the word I'm looking for but like a sense of unease Because at the end of the
day If You're looking for the Easy way out there isn't we need to try it out we need to actually Sit and do the work um and and that would be I think my My opening Part as then we need to and and This is what I hope to accomplish um this afternoon right in the Hour and a Half that We have move at least a little bit to in the right Direction Thank you Thank you so much phip this was very illustrative and a very good way to Start I Guess it puts Us to
initial concepts as we discuss this topic One of them has to do with the idea of ethics versus Human Rights and ethics Being Something that has a lot of contextual a strong contextual Element Human Rights are the standards That were already globally agreed and that we try to Stand for and the difficulties of working in in an environment based on ethics Because they Will var culturally they Will var within Time As you saying and We Will have a lot of Uh challenges there and also the idea of the sandboxes one part of the idea at
least as a possibility for Us to Experiment and to test some hypotheses I Guess I will Leave it there The way you put it Because I Guess others Will Touch on these ideas too from different perspectives Because I know that the idea of regulatory sandboxes and How it is Being Advanced for Ai at least in different Uh regions of the world is Still Pretty exploratory also right so part of This conversation is also touching on How do we better Define Those How do we Define the key Objectives of Those In A Way that they can
actually allow this learning process in this contextual process I think This is very interesting well I'll pass right now to Jonathan Mendoza Jonathan is a doctor of Law and Master in general Data protection regulation of the european Union he is currently the secretary of personal Data protection at inai Uh in Mexico Where he indust standards guidelines andom Privaban you and please feel free To Do your intervention in English or in Spanish slowly spanish so brazilians follow up with you Thank you so much Good afternoon buenas tardes dice yval no arari que lo primero que tienes
que saber en un sistema de Inteligencia artificial es si estás hablando con un humano o con una Inteligencia artificial ese Debería ser El Punto de partida Me Parece que los sandboxes regulatorios son una herramienta muy importante para poder definir Cuáles son las prioridades legislativas que tiene que haber respecto al uso de tecnologías disruptivas Hoy estamos hablando Inteligencia artificial generativa pero ese es el el trending topic No pero en 5 años quizá hablemos de computación cuántica No lo sé o de alguna otra tecnología hace 5 años hablamos de Big Data Cómo tenemos que Regular la tecnología
Cuál es el mecanismo idóneo para hacerlo porque hay un Boom a nivel latinoamericano y a nivel Global en general por regular los temas de Inteligencia artificial en mi país por lo menos hay alrededor de 30 iniciativas que tienen que ver con regular temas de Inteligencia artificial están interviniendo todos los actores en la en la en el diseño y la elaboración de esta regulación o no y cuáles son los efectos positivos o negativos que puede Tener implementar una norma para la cual una sociedad no está preparada hablaban inicialmente en este panel de ver modelos que son
funcionales en otras regiones o en otros latitudes e incorporarlas a sistemas jurídicos de nuestros países si quieren hablar de soberanía de datos quizá podemos hablar de de China como un Punto de partida O si quieren hablar de una norma que es fuerte en cuestiones de protección de datos personal persones quizá Hablaríamos del reglamento general de protección de datos de la Unión Europea del gdpr Pero eso funciona en otras jurisdicciones O no Cuáles son los mecanismos que tienen los gobiernos para poder adoptar este tipo de normas si no corresponden a sus cuestiones eh sociales políticas económicas
culturales por muy bueno que sea el estándar si la norma no se adapta a la realidad de un país pues está un poco e destinada al fracaso es por eso que los Open Loop los Sandboxes son herramientas que son muy útiles y funcionales para identificar en un espacio seguro y que es Temporal y por ahí hay una guía de orientación de la oe y ya hay una implementación en varios países en Francia en Reino Unido Brasil tiene justamente un sandbox abierto para temas de consulta pública Singapur lo hizo en Inteligencia artificial generativa con estos resultados
puedes tener un mapa o un panorama general de cómo está el tema en El país y entonces sobre eso quizá ir a mecanismos de soft Law y posteriormente a regular Porque si regulas sin conocer el contexto eh Tienes un riesgo muy alto de equivocarte Y si esa Norma no es útil Entonces pues no sirve para los fines que está persiguiendo en realidad o sea se vuelve letra muerta es algo que es muy difícil de lograr o de implementar existe el derecho positivo se tiene que aplicar pero su aplicación es de muy Difícil eh concreción Y
entonces los responsables terminan en muchas ocasiones por no aplicar las normas sino por ver eh algunos mecanismos alternos para poder solucionar el tema porque eh la tecnología y la innovación no va a parar no se va a detener y menos por un tema de de regulación o sea más bien bajo una óptica de eh de de defensa de libertades y derechos fundamentales Cómo esas regulaciones pueden ser útiles para impulsar temas de economía digital de Comercio electrónico y Y cómo el entorno o el ecosistema se vuelve eh positivo en todos los sentidos y no está un
poco fracturado o digamos dividido como podría suceder si no intervienen todos los actores en la en la elaboración de las normas hasta aquí Me quedaría yamil en mi primera intervención Muchas gracias excelente Jonathan Muchas gracias y creo que nos traes otra reflexión una reflexión Complementaria English foll the Flow gu addal consideration the idea of regulatory Sand Which has to do and How do we test models within this controlled environment and Now we were hearing from you as part of the Data protection Authority in Mexico How these instruments can also assist on regulatory innovation And I
Guess There is a question around What do we Understand as innovation and What are the criteria or the the limits Under Which We want to understand innovation in both Ways What extent Can we innovate at the legislative Level or the the normative Level and to What extent do We want to allow this Tech innovation in a control setting You are saying hopefully We Are Not Talking About life and Death situations around innovation What are other criteria we can think of If We think of these as instruments that are Also grounded in in Human Rights I
Guess there are some very interesting Uh insights for Us to continue thinking about Now like to move to Our online participant Uh cis de la pea She has More Than 20 years of experience working on public policy and regulation of digital technologies She Now advises The mexican Congress as a founding member of the National Alliance for artificial intelligence anya and recently She coauthor the proposal for a National Ai Agenda currently She serves as the International affairs director at the Mexican Academy for Cyber Security and Law c I hope you can hear Us please please feel
free To Do your intervention in English or in Spanish As you prefer you have Seven minutes for this First intervention Thank you very muchila and Thank you very much for the invitation Uh I think I'm also going to move to spanish If That's okay for viw um and and yeah right listening at my prev Interv spanish spanish escuchando a mis colegas con sus intervenciones previas hablando de Cuál es esta Barrera entre ética derechos humanos y Y estos Marcos regulatorios que deben ser ajustados a lo que es el componente cultural eh jurídico social y legal de
cada país e coincido en que este mecanismo sandbox es una manera de innovar también en el componente regulatorio eh justamente eh ahora que Tenemos la Inteligencia artificial sobre todo en el componente de tecnología el componente regulatorio siempre ha sido eh ambiguo en el sentido de que los reguladores o los que están por lo general encargados de regular no siempre han sido definidos con Esa visión per sé dado que las e las tecnologías van cambiando eh continuamente entonces lo que lo que sucede en esta coordinación institucional en el componente digital dado que el componente digital es
la Sombrilla de todo lo que sucede en todos los otros sectores e ha sido ha tenido este entendimiento un poco eh con una con una velocidad más retrasada en cuanto a cómo se van a ir regulando las nuevas los nuevos modelos de negocios o los nuevos emprendedurismo el nuevo emprendedurismo que suceda y ahí entonces empieza Una descoordinación institucional entre Quiénes deberían regular y quiénes deberían tener una opinión también como lo es la sociedad Civil para proteger los Derechos Humanos e el el regulador del sector per s Hablemos del regulador en el sector salud o en
el sector energético y el regulador eh Y los reguladores que tienen que ver componentes específicos en el componente de tecnología y justamente esta complejidad institucional que puede presentarse en cada país y que desde el punto de vista de países en Latinoamérica no siempre es tan claro Eh coincido con mi colega Jonathan en que copiar y pegar normativas o regulaciones es la la receta para el peor desastre porque justamente estas normas o estas regulaciones no se ajustan a los marcos legales y jurídicos de cada país en ese sentido creo que los sambox regulatorios son también una
nueva forma de entender o nos dan la oportunidad de entender cómo los diferentes actores tienen que ir tomando un rol específico En ese en ese proceso de regulación y estoy hablando de la comunidad técnica La Academia la civil y los el sector privado y los actores e del sector gobierno perc los reguladores perc eh este proceso de innovar en Jonathan mencionaba Cuál ha sido la experiencia de México y creo que México eh justamente se ha enfocado mucho los últimos 6 años se enfocó mucho a entregar el componente regulatorio al Congreso eh justamente creo que desde
el Gobierno Federal quien hace política pública eh Por ahí no quedó tan claro como la agenda en el componente digital iba a quedar implementada en materia regulatoria y mucho menos en una agenda de Inteligencia artificial donde todo se exacerba y donde todas las e pues las acciones que hay que tener para cuidar los Eh pues para salvaguardar eh los riesgos a los que están enfrentando dados los usuarios pues tendrían que Estar atendidos Y entonces esto se se se en México esto se se arrojó o se cayó sobre todo en el en el congreso y Jonathan
lo dice muy bien más de 30 iniciativas con un componente de alguna u otra forma en materia de de Inteligencia artificial que no están coordinadas y muchas de ellas no están ni siquiera basadas en la suficiente evidencia ci entonces volviendo a lo que es el sbox regulatorio como un mecanismo innovador De participación de todos los actores involucrados no solamente permite dar un proceso de prueba eh controlada para ver cómo funciona esta nueva tecnología en x o o y sector sino que también otorga a los gobiernos una una visión o una oportunidad de repensar cuál es
esa coordinación institucional que debe suceder entre todos los actores involucrados a la Al momento de regular de acuerdo a las capacidades y a las oportunidades que tú tienes en tu país O Sea culturales normativas legales jurídicas y de enforcement también e mucho de los fracasos de este proceso de copiar y pegar normativas de otras regiones para latinoamérica es que en otras regiones se Norma o se legisla se regula de X manera porque hay enforcement pero en muchos de nuestros nuos países no hay enforcement eh Y con el riesgo que eso que eso implica para los
usuarios entonces ahí quiero yo dejar mi intervención Eh las ventajas de un mecanismo sandbox eh van mucho más allá de solamente a implementar La regulación sino todos los actores involucrados eh aprenden del proceso se van eh Van Van Van aprendiendo van eh van adentrándose más en el proceso de lo que es la tecnología con los derechos humanos y viceversa con la academia Pero también es un proceso de oportunidad para los gobiernos en aprender como esta coordinación institucional se va a ir dando en este Componente tecnológico Thank you so much c this was an Amazing intervention
I Guess Because it allows Us to deum that you talked about Uh participation multi stakeholder participation you talked about um shared learning processes and you also talked about Or You Leave Us with the reflection about accountability and transparency also within Those process right allowing Us to go a bit further as we identify that Our governments or several of Our governments are seeing this as a key instrument Uh Which I'm understanding that from your perspective is It's an interesting Step towards Uh finding regulation that is contextualized regulation that is evidence based but How do we actually
design Those Those instruments It's very Uh is a very important Element And I Guess you bring Us a lot of insight to Continue on Our discussion I'll pass Now to lorine porul She is the executive director of the Data sphere initiative Prior to leading the Data sphere initiativ She worked as oecd the atu among other organizations lorine is an affiliate to the berkman kin center for internet and society at Harvard University University Where She conducts research on Data for development Please lor Thank you so much jamila um and It's a pleasure to be organizing this
Uh this Event with you today in cpd patan Uh It's very important to talk about sandboxes in a context um of people that are working around Data protection Uh in the region and before I go to talk to sandboxes I Want To talk about complexity and the insufficiency of institutional arrangements Uh and legal arrangements that We have today and The Reason Why complexity is important um also to do a call back to Our name the Data sphere Uh is a concept that we put For to understand this as a New space of action right the
Data sphere is conceived Of Not only the sum of all the Data that there Is there and some that includes not only personal Data but a lot of non-personal Data in terms of volume Uh the majority of Data Being created is by sensors metadata so there's a lot there in terms of Data that is not personal um and the barriers That are very fuzzy Uh in between so Data say public and private norms but code as a norm in itself and the third Element is the notion of human groups in terms of the different public
and private human groups so within this TR of elements of Data sets norms and human groups we find ourselves in New sphere of policy action of economic action of social interventions as well and It's very Complex it goes Beyond uh uh the Fragmentation That We have within sectors and policy areas and It makes it very hard for Us to follow that from a regulatory perspective or policy perspective or social civil society perspective as well and observing All of that is important and naming it as a New space is important as well Because it it needs
to make it Clear for Us that it's we need New Tools the way that Our institutions have been developed It's very much vertical if you look into Traditional sectors be finance or Telecom Those are very defined entities that were regulated and we could actually Point them into the value Chain and What Services they were Being they were producing and also in terms of borders right they Were they they Had their Services defined within a certain geography and right now It's a whole different World and we don't have the Tools and the institutions to respond to
that and there's a a very established Field already Looking into complexity and complexity studies and What is observed and there's many different um I mean definitions um there's a long definition into What a Complex adaptive System is but typically What you do is that you need to observe and adapt to that it's very hard for you to model complexity so that means that um You Can't really get linear approaches to How you respond to it you need to be very flexible and adaptive um and That's Go that goes through the studies of certain molecules all
the way to the to to studying How Birds move so Now I've taken you to a Journey around complexity How to respond to that in a World that you know the borders of different topics and Services are so fuzzy we need we need Tools that are more agile and This is Where Sand boxes come in Sand boxes um clearly Uh a Word and a concept that has been put in a lot of different interventions Um and but What's interesting about them is that they they they from different innovation labs or innovative Uh spaces that have
been already implemented within the public sector Because They are necessarily As you said They are temporarily bound so you actually have a defined space Where the experiment happens and you Always include the private sector Uh In A Way so It's not only innovation for the purpose of the public sector innovating its own Services But You actually invite different stakeholders to that that experiment and What's interesting sandboxes also is that you You are observing something and integrating on it traditional Ways of doing regulation they they Uh rely on a very linear approach you produce um um
an input for a consultation and a study and then you develop your First Draft you do consultation you review that you approve it in the end regulate and forget What Sandboxes and regulatory sandboxes in particular could Uh allow is for a Loop the same way that we think about software development When we think about versioning of that software we could think about Looking to Our regulate regulatory frameworks as well in terms of loops and learning and When we deal to the complexity and the Speed and and and the velocity of How chang is happening we
need New Tools sandboxes can be that but What's It's not not all Rosy Uh we need to think about How to improve sandboxes in the way that it's Uh uh It's actually Being developed particularly Because we need this to be not only responsive to the different context Uh local National and regional but also considering How how we measure success how we deal with failure how we actually use this to coordinate between different stakeholders um right now what I see is a tendency also to somehow Restrict the scope of sandboxes um it cannot only be about
Looking towards only one lenses If You're Looking into Uh Ai sandboxes it can only be about One lenses about Data protection there's so many different dimensions related to Ai that Don't relate to personal Data How do we deal with that right it can not only be through the lenses of either regulatory sandboxes In itself The way we Define is that sandboxes can be in two archetypes either regulatory or Operational and a hybrid of the two as well and by operational we mean um Data spaces or Data Lakes spaces Where You're actually dealing with Data in
itself and operationalizing it and there are a number of different Sand boxes that are regulatory and that Are Not necessarily lifting regulation itself right and so there are a lot of preconceptions and assumptions within the World of sandboxes that need to be matured and developed and That's work that We are Doing Now dat initiative um We have put Uh uh and published a report In 2022 which Was the First One to look into the role of sandboxes for Data and particularly the potential for crossborder sandboxes and since then with been developing projects in in number
of different spaces in the world in africa more and more rec countries in latin America as well and also in Touch with many different countries in Asia and europe trying to understand what are Those good practices What are the spaces Where more improvement needs to be done and How can we actually be more structured about thinking about sandboxes and applying it for other sectors and also ambitious to think about How this could be a space Uh to also propose a New type of International cooperation Uh Because so many of Those issues Are Not bound within
countries we need to be able To talk to each other and share good Practices I've already talked to much I'm going to stop here and continue and continue later Thank you this was excellent Thank you so much lorine I Guess you also bring other elements for Us to think of this um instrument for instance the different types of different formats or different objectives they might have And I Guess something that comes to my mind for Our Uh q&a in the end It's should we think of contextualizing Sand boxes also as a Tool for Our particular
um institutional settings institutional arrangements as cc Was also mentioning and How do we think of this complexity in terms of the different authorities that should or could be involved in these initiatives and How they Will participate in the cases that we seen in the region I will pass Now to Lucia Camacho for her intervention Lucia is my colleague os digitales Where She is a policy coordinator She is a lawyer specialized In public policy and She has a Masters in Human Rights Lucia is the author of Our paper that is to be published about sandboxes And
I hope She can even bring more insight into conversation right now I'm sure She will be able to to Us more ideas ands convers thank luí thank jamila continue in Spanish joining and Jonathan de qué estamos hablando cuando hablamos de sandboxes y sandboxes regulatorios porque no sé si es una idea Que todos tengamos Clara varios elementos de esta figura han sido aportados Por quienes están en este panel no una figura que permite a los reguladores aprender los problemas no los retos que traen consigo las innovaciones particularmente las innovaciones tecnológicas para aprender de ellas en la
práctica y luego entender cómo regularlas no es como la definición Al Menos que aporta La Academia y luego cómo eso se pone en la práctica ahí las aproximaciones varían Según a quién le preguntemos no en el estudio que hicimos en derechos digitales miramos particularmente la experiencia en América Latina porque vemos Que no hay una visión uniforme sobre esta figura que en la academia parece digamos haber mucha mayor Claridad sobre Qué es O al menos cómo funciona en nuestra región a propósito de las diversas estrategias Para la Inteligencia artificial en países como República Dominicana chile Argentina
Perú Colombia Brasil entre otros se ha anunciado la adopción de esta figura como una entre tantas otras que está en la caja de herramientas de los reguladores y que los reguladores consideran va a ser la que permita responder a este reto que trae con consigo particularmente la Inteligencia artificial en donde cualquier Norma que se expida ya adolece De una suerte de digamos estrechez por la naturaleza misma de de la tecnología luego en nuestra región particularmente hay dos experiencias que nos permite entender cuán diverso puede ser el entendimiento de los sos sandboxes regulatorios en la práctica
particularmente el caso colombiano y el de Brasil aportan digamos bastante a esta discusión sobre todo el de Colombia mi país eh es citado como un caso de éxito que quizá merezca una visión mucho Más crítica la autoridad de protección de datos en 2021 eh desplegó el sandbox de privacidad por diseño y por defecto aplicado a la Inteligencia artificial y más o menos proponía lo siguiente a todos los innovadores que tuviesen una Inteligencia artificial les invitaba a concurrir en un espacio de prueba supervisado por la autoridad para que la autoridad les dijese Cómo Aplicar el marco
regulatorio vigente en caso de que hubiese dudas la autoridad no prometía como suele hacerlo la figura del sandbox regulatorio en la visión más académica la autoridad no prometía excepciones regulatorias qué es lo que permite a las innovaciones tener esta suerte de espacio libre de juego y testeo sin La amenaza del regulador y eventuales sanciones la autoridad colombiana No prometía nada de ese estilo lo que hacía era advertir en Cambio juegue testé yo lo supervis pero juegue con lo que hay con la ley existente yo le asesoro no le voy a prometer que no lo voy
a sancionar en el futuro puede que lo haga no la asesoría que le voy a dar no es garantía de nada Si se porta mal con seguridad lo voy a investigar luego esa experiencia cuando vemos los resultados no porque nos prometen los sbox es bueno el regulador aprende luego entiende Cómo regular no todo bien luego Cuando analizamos críticamente los resultados de los sandboxes los dos despliegues que hubo en 2022 y 2023 entendemos que los innovadores que acuden a este tipo de entornos de experimentación segura No tienen esa digamos no ponen en el fuego innovaciones que
generen preguntas serias en materia de Derechos Humanos la mayoría se trataba de s de innovaciones en Inteligencia artificial con chatbots Para asesorar en marketing a consumidores para responder preguntas no y dar respuestas digamos una Innovación que no Debería ser problemática al menos no es el tipo de innovación que uno esperaría que acudiese a un sandbox e y luego de 2023 deja de existir deja de ponerse en prueba esta herramienta no entendemos muy bien por qué pero deja digamos abandona la autoridad esta herramienta que puso en práctica y no entendemos luego Cuáles fueron los Aprendizajes de
ese acompañamiento que nuevamente no prometía excepciones simplemente le decía al innovador aplique lo que hay yo le digo Qué hacer en caso de dudas una suerte como de compliance tercerizado en la propia autoridad no donde la asesoría la da el mismo regulador luego la experiencia de Brasil que no se ha puesto en práctica pero que promete cosas interesantes la autoridad en Brasil dice yo quiero testear mi propia Regulación particularmente donde hay vacíos porque la Ley General de protección de datos desde luego no respondía en su momento porque no podía hacerlo a los retos que traería
consigo la Inteligencia artificial generativa y decía la autoridad o dice a la autoridad en este documento de consulta que diseñó para poner en prueba los sandboxes quiero testear el artículo 20 quiero ponerlo a prueba el artículo 20 sobre transparencia transparencia ver cómo Entendemos esto de la explicaba en la guía porque la ley nuestra no tiene nada en es es sentido luego advierte nuevamente la autoridad y esto quizás sea como una señal de cautela de las autoridades advierte no voy a dar tampoco excepciones no voy a prometer que no voy a sancionar a quien pueda cometer
un error en el entorno de prueba y fuera de él luego cuando despliegue no a los consumidores esa innovación tecnológica Arriesgada no voy a conceder excepciones pero sí que voy a dar una asesoría más o menos un enfoque parecido el de la autoridad en Colombia luego entonces uno queda reflexionando sobre cuál puede ser la promesa en la práctica de una herramienta que al menos desde la visión de académica debería generar flexibilidades normativas y regulatorias no para permitir a las innovaciones no desplegarse en un entorno supervisado controlado temporal pero sin La promesa De una eventual sanción
para que los los innovadores no se sientan amenazados no luego Esta aproximación cautelosa no es puramente exclusiva quizá de el enfoque latinoamericano que parece estar emergiendo sino también lo pudimos ver en experiencias en en Europa que que también pusimos bajo la lupa para entender si aquí también como en otros herramientas regulatorias hay una suerte de de efecto Bruselas y también esta suerte de Cautela donde no hay promesas arriesgadas por parte de las autoridades pero las autoridades sí quieren entender qué está pasando no con tecnologías que no terminan de comprender del todo eh se comparte también
en Europa en la experiencia eh particularmente de Francia que es donde ha habido sandboxes también en Inteligencia artificial y protección de datos luego solo para cerrar dejaría como esta idea de hay una suerte de brecha entre lo que parece Prometer est figura y lo que en la práctica hemos visto al menos de los despliegues no que tenemos en nuestra región Porque todos los demás países tienen promesas ideas el caso de Colombia y El de Brasil que son los más avanzados nos permiten entender que las autoridades quizá También tienen sus dudas Quizá no no las han
manifestado tan claramente pero sus aproximaciones más cautelosas sin prometer no excepciones en materia de derechos Fundamentales particularmente yo creo que nos nos deja algo Quizá para continuar la conversación ahora Thank you so much Lucia And I Guess the the to see how This is already Being implemented is very valuable as we continue to think of criteria how this could be improved What is the actual use of these types of of mechanisms I will Start the next round of interventions Imagine you have a lot of things to react to each other and to my my own
Provocations during this conversation I will Leave some questions some guiding questions that we prepared for the session But You please feel free to react to One another or to identify a question that resemble With um Your thoughts and experience the questions That We have prepared are the following one is what are the promises and advantages that regulatory experimentation with sandboxes to the discussion on Data governance and Ai governance for latin America What lessons or insights do the design or deployment of sandboxes for Ai in the region offer and What risks can they represent for the
protection of individuals How does the Complex complexity Element of Ai impacts or complicates the regulatory Over site by a single National sandbox operation And I Guess several of the the speakers touched that Point on the institutional complexity and also the Jurisdictional Element that lorine mentioned and is very relevant here Uh and The Last question What sort of International collaboration could help Us address key challenges and risks Uh each Of You have five minutes for their intervention and I'll follow the same order of the First round so please Philipp I was hoping I get to be
First Again I was I was Crazy you know I was dying to react Uh to some of your points um I do have comments on your latest two Questions but before I get there I want to build from Where you stopped um if you allow me um but Bear with me Uh I'm going to make a big turn but but I promise I'll get to my Point at the end so We have been humans for at least A few thousand years I think we can all agree on that right So The First time we Had
a man-made existential threat for humankind was 80 years ago there's there's an interesting movie about that It's an osc movie in Case you haven't Seen it I think they call it openheimer um but nuclear Science is also really Great for humanity so it brings Us more precise archaeology it brings Us Uh nuclear medicine it brings the light and the Energy that Lights Our Cities um and we Are all very excited about Ai right so We want to get to work We want to get to exper We have all Crazy good startups coming And I can
say Because I am an investor to many of them um but I honestly don't envy the work of Regulators um they're really overwhelmed um and It's a really ungrateful Job Because we don't remember When they do a good Job and and you know Nothing happens as as It's supposed to but we do have a very long Memory of When they make mistakes so um quick survey Do you know anything about the locomotive from the 19th century all right It's One of my favorite examples so um in in late 19 Century the British parliament approved a
couple of laws that they then you know Removed later on um that basically in summary um created a scenario by Which if you Had Any um mechanized vehicle Walking in the streets Of You know anywhere in the UK you needed Somebody to be Walking of that vehicle waving a red flag um and this Was the purpose of that was to alert um pedestrians that a vehicle was coming Which you know makes some sense right but in retrospect it seems very silly and this Of course hampered the development of engines for At least 30 years Because
Cars or locomotives or Whatever they couldn't Go past Five or 6 ket per Hour Because That's the you know a decent Walking Pace for most people um and so When we Look at regulation Uh for for something that is emerging we need to think is this going to be the Next locomotive act Right um And My provocation here Is there are two scenarios for Ai regulation If We Don't do sandboxes the First is we don't regulate at all and Many people don't want that um and the other Part is We have a locomotive act but
Of course We are only Going To Know that many years in the future and the Thing about this vis locomotives and Cars Back In The late 19th Century is that Because the barrier of entry is so much lower and the proof of that is the amount of startups that we see um the risk of falling behind is much much higher and the cost of falling behind Not Just as as a you Know Group of companies as a Group of individuals but as as a nation as society It's much much higher so We are probably in
the First time in the past thousands of years Where with marginal effort compared to the last you know especially Last hundreds of years with marginal effort a nation can actually very quickly get ahead on the on the innovation on the economic Race um This is not Just about creating code of course There is a bunch of other myriad Of things that Go into that scenario but effectively is a much lower barrier than we've seen in the past um and of course we're Looking at many challenges that Whatever sandboxes we come up with Because of course
If it If it isn't Clear I am a proponent of Us doing sandboxes we need to Look at anything from technical Security to privacy to unemployment to purpose to addiction to solitude um and All of the questions That Not Just economical but as as human Beings we need to Face When we're Looking at the Boom of Ai and amidst of that Of course we need to Start thinking about Data governance we need to Start Start thinking about upskilling about reskilling we need to Start thinking about the right um environment for for flourishing of New businesses
we need to Start thinking about investments we need to Start thinking about partnerships Because um It's Not Just about the code as I mentioned but we need Data centers We need gpus we need Clean Electricity to make All of that happen right so even though the barrier Is lower there are Still many things that we need to get right um and so my question and This is to Finish Where I promis That I be um Getting to your Point is Where do We want latin America to be in 20 years from Now um And I
think This is the time Where we need to choose If we're going to have the locomotives with people waving red Flags or If we're Going To Have The flying Cars of the jetsons right um and to go to your point to your Point How much time do I have I have No time um so Let me stop here and I'll get quién en su sano juicio se va a acercar a una autoridad que posteriormente te va a sancionar con la información que le estás proporcionando o sea es ilógico esto tiene que ver con perf las
autoridades todas las autoridades los funcionarios públicos los gobiernos parecería que este tema de Defender derechos no tiene que ver con dar una solución integral sino con que haya consecuencias para quien lo hace mal sin entender que muchos más potencialmente lo pueden estar haciendo mal y que esa autoridad ni siquiera tiene conocimiento de esa situación Entonces el perfil de las autoridades por lo menos en materia de protección de datos personales que es de lo que yo puedo hablar porque un poco conozco más eh No tiene que ser eh Persecutorio o o de ir hacia sancionar eh
empresas compañías eh instituciones porque cuando hacemos eso es que ya se transgredió un principio ya no se respetó un deber ya se violó un derecho Esa es la parte coactiva esa es la parte reactiva esa es la parte opuesta a los sandboxes que son para a planear eh Y y voy hacia ese tema me parece que es una buena herramienta para sentar a todos los actores en la mesa en un espacio que no sea controlado Por alguno de esos Actores Y esto es importante porque Generalmente alguno de los actores eh por por costumbre o por
tradición siempre el sector público o los gobiernos o las autoridades buscan liderar y En consecuencia de una otra manera controlar las conversaciones es decir yo te invito a mi casa para que me digas qué es lo que estás haciendo te esté aquí veamos los usos de la Inteligencia artificial y si está mal no te aseguro que no va a haber una Consecuencia es eso no es tener una silla en la mesa de conversación las autoridades están llegando o estamos llegando tarde a las conversaciones por es todo el actuar que tiene que ver con nosotros es
reactivo no es en un esquema de gobernanza gobernanza de datos personales gobernanza de la Inteligencia artificial gobernanza de la tecnología disruptiva que ustedes digan hacia adelante las normas siempre van a ser estáticas siempre van a ir atrás de la Innovación y la tecnología y la innovación y la tecnología no dejará de existir o de avanzar en función de la Norma que se aprueba en el país que ustedes gusten y manden O sea que si mañana hay una regulación de Inteligencia artificial en Brasil no es que van a dejar de desarrollar verdad o ya no van
a dejar de entrenar sistemas de Inteligencia artificial o no va a haber desarrollo de algoritmo no como e los legisladores los actores Relevantes los tomadores de decisiones se involucran en una conversación en una situación que ya existe el día de hoy cómo dejamos de generar una bola de nieve que es un círculo no virtuoso que es negativo y formamos parte de las conversaciones desde la desde una posición que no sea burocrática tenemos que entender con flexibilidad Cuál es eh el rol que tienen que tener las autoridades en estas conversaciones la protección de Datos personales no
puede ser una barrera una obstrucción al tema economía digital de comercio electrónico ni al desarrollo global de esto O sea la Inteligencia artificial va a sumar 15.7 billones de dólares para 2030 no va a parar por ninguna regulación Los costos los van a asumir si es que hay una contingencia por no atender una Norma creo que estamos en un momento importante en Latinoamérica para poder entender esto y cambiar el rol que Tienen las instituciones y las autoridades eh No no lo ha dicho sisí seguramente Ahorita lo va a decir pero en en México hicimos un
ejercicio a través del senado de la República y la senadora Alejandra lagunes que es una alianza Nacional de Inteligencia artificial y de ahí derivó un sandbox justo con la embajada británica a ver nos tenemos que Escuchar todos no no hay una sola persona que tenga e digamos eh la respuesta correcta y Única por eso es multiactor por eso es multidisciplinario y generar esos espacios lo que nos va a permitir actualmente es que el ecosistema sea más seguro que de mayores certezas hacia los usuarios y que los desequilibrios existentes al día de hoy entre servicios digitales
y usuarios se vayan balanceando en función de las decisiones que se tomen responsablemente por todos los involucrados Y qué se trata de hacer con La regulación y con eso concluyo Como hay una simetría en en internet o en los ambientes digitales se trata de balancear con La regulación esta asimetría pero el problema no está ahí el problema es cómo cambia esta conformación monetizar datos discúlpenme ustedes no está prohibido si hay una causa de legitimación de por medio es posible hacerlo El problema es que del otro lado no hay conocimiento No hay autodeterminación informativa no se
ha socializado el derecho no se ejerce no Se consolida una cultura a La regulación tienen que ir aparejadas otras acciones como política pública como la implementación transversal en las los gobiernos o en las instituciones hasta ahí me quedo ahora jamila pero creo que vale la pena reflexionar en torno al papel que tiene que tener cada uno de los gobiernos en estos temas muchas gracias voyar de lo que decía Jonathan y sobre esta verticalidad que hablaba lorraine acerca de cómo está Esta estructura institucional en los países en Latinoamérica e eh Jonathan nos decía hay siempre quiere
haber en este contexto multiactor que es lo que nos ofrece los sandboxes eh esta posibilidad de trabajar en un contexto m multiactor siempre habrá por ejemplo alguien que quiera controlar y por lo general eh el gobierno cree que es el rol de ellos es tomar ese tomar ese papel e coincido mucho con lo que dice Jonathan completamente esto debe ser en Un componente multiactor Y entonces hablando específicamente ya que Jonathan me dio oportunidad e de hablar de este proceso de análisis de sandbox regulatorio que se hizo en México fue Prácticamente explorar la viabilidad y ver
si estábamos en la posibilidad de implementar un sandbox por lo que mencionaban mis colegas si se entiende de la misma manera en todos los actores Y saben primero saben o conocen que es la figura del sandbox si se podría Implementar en un contexto de coordinación institucional es decir nosotros en este proceso de análisis para la el apetito de implementar un sbox reunimos a los principales reguladores en el contexto digital el inai la cofes que es el el regulador en materia de competencia y el Instituto Instituto Federal de telecomunic iones junto con con con la autoridad
de protección de datos y el senado y un caso interesante o dos resultados que yo Quiero destacar de este análisis fue primero todos los participantes incluyendo sociedad civil Academia y industria privada coincidieron en que sí debería haber un sandbox regulatorio en México pero no solamente para Inteligencia artificial sino para cualquier tecnología y en cualquier seor Y eso daba entonces un poco la oportunidad de repensar o de de conversar acerca de cómo esta coordinación de los reguladores debería Darse los que regulan eh la normativa en materia de datos en materia de competencia en materia de telecomunicaciones
y después los reguladores sectoriales per s el regulador sectorial en materia de energía por ejemplo o de Educación o de salud ustedes nmbr eh y en ese sentido entonces otra de las conclusiones de este proceso fue e sería ideal tener un una especie de digital coordination regulation Forum un poco Como para justamente atender la problemática o el reto que mencionaba Jonathan de no tener un solo regulador un solo ente que tomara el control de ese proceso regulatorio ya hay ejemplos como ejemplos de ese de ese tipo de organización Inglaterra Canadá va e Australia tienen proces
tienen eh organizaciones similares e y uno de los retos que yo veo sobre todo en países en América Latina justamente y vuelvo creo que a donde Empecé mi intervención es precisamente eh A esa complejidad institucional e al final del proyecto eh algo que fue muy interesante y yo estoy muy contenta con los resultados de ese proyecto fue tener sentados a los tres reguladores eh incluyendo el senado a los cuatro reguladores eh Como paralelos o como eh periféricos que deberían estar hablando de de de cómo Apoyar el proceso regulatorio en un proceso sandbox E sin embargo
lo que sucede en México es un proceso Pues igual que en muchos países de latinoamérica la falta de continuidad de políticas públicas y el cambio de la estructura gubernamental en cada cambio de gobierno complica mucho eh Eh pues eh seguir implementando estos debates Y estos diálogos y llevarlos a la práctica probablemente México ahora no sea la excepción eh esperemos que los cambios en el cambio en gobierno permita eh tener una continuidad en estos en las Propuestas que han surgido de estos diálogos pero ahí vuelvo entonces a la complejidad no de implementar el proyecto un mecanismo
sbox per s sino la complejidad de la que hablaba lorin de tener esta coordinación institucional que contempla esta complejidad de conversar en un ambiente multiactor para lograr resultados específicos y creo que ahí es donde yo enfocaría más las lecciones que hemos aprendido desde desde México si bien hay mucho apetito Por el ecosistema digital para implementar un mecanismo sandbox como un mecanismo innovador de regulación en materia tecnológica con lo que tenemos que enfrentarnos también es con esta complejidad en la verticalidad de la institucionalidad del gobierno y creo que esos son retos que deberíamos de tener Pues
un poco ya eh quizás más atendidos o solucionados y que es por eso que no podemos copiar y pegar Regulaciones de otras regiones donde A lo mejor Esa complejidad institucional está un poco más resuelta creo que ahí yo dejaría mi intervención Thank you c and immedi toor Thank you so much and I'm bu several Those points and to answer your First the First phrase of of your intervention Jonathan Which was Who in their s Mind as the private sector would Come To a regulator and and Say You know exactly What their problems are um and
and It's interesting Because we've been working very closely to the singaporeans and they have implemented so many Times their sandbox that talking to the Head of the of the regulator Uh Im they Us that Listen right now It's the private sector their compliance Team that comes to Us to propose sandboxes and for me that answers the potential for sandboxes and what they can be and for me the key Word is Trust I mean That's basic prisoner dilem Theory if you Run one problem once you have One which is easily the whole society loses out right
there's No Trust between the stakeholders you need to Run the same experiment several Times for people to understand and to be able to anticipate What's going to be each other's behavior and build that Trust That's What for me the potential sandboxes are and and we need to be able to be intentional about This objective right and It's Not Just playing around with rules and Uh and you know pretending that we lif this or leing that and we need to be intentional about ultimately We want to improve people's lives and make sure that We are saving
Our Planet right and and How do we do that How are we able to build better relations between public and the private sector and all stakeholders affected and and sandboxes is one of the Tools It's not the only one but Certainly is one that responds to complexity and response to that need of actually be more interative In A Way that we're approaching this and Uh and for me the examples of singapore Is inspiring although You Know It Is A different Level of complexity institution they've been doing that Uh they have a very Close integration with
the private sector in several the policy and regulations that they do so There is a very specific cultural context there But I do think that it it gives a gives a lesson in terms of What sandboxes can bring to the table and for me the keyword There is how do we allow this to be a space Where we're building Trust and that Uh to your Second question is what are the lessons right And I think That's why we Need More evidence um in terms of What are the good practices for sandboxes and How can we
do this kind of study and mapping and and doing an x-ray of the different sandbox experiences to Understand what was their actual result and What We have seen And I mean we have a large inventory of different sandboxes experiences around the world not only around Ai but on different Uh sectors on Ai We have identified Almost 20 different sandboxes today but What's interesting here is that most of that information is not public right and If you've done research in the region You See That as well and there's very Little information online so How Can we
change That How can We have Uh a structured approach to how we are sharing information around sandboxes How can we make this part of good practices when you actually develop and design your sandbox that you are sure that You're Going To report It Is there a Forum Where you actually reporting to right now in the context of the Ai act establishing designing a sandbox is not um just a Nice Thing to have It's a requirement right All the countries Within the eu Will have to um develop their own Ai sandbox and report on it to
the european commission Uh to the Ai Authority What's What's going to happen in latin America then How are we Going To coordinate This Is there a space Where we actually can share Those experiences and good practices and build on each others to understand How to set up better governance for sandboxes themselves How to actually create spaces Where you are consulting on sandbox Where you you have an assessment framework for sandbox itself How are you mitigating risks within sandbox as well and What happens Uh When it fails everyone's very happy When sandboxes Go right right um
I mean auditors Are Not very happy When it doesn't and so so who's going to be accountable for that and how we going to build lessons learned from Those failures as well that can be very informative and and for me that needs to be more structured and That's why we're in a Mission to actually build that space Uh next week we're going to have the inaugural meeting of the global sandboxes Forum um and Uh and What this space aims to to create It's It's a Community of practice of of practitioners around sandboxes is a space
Where we can share resources capacity building good practices around sandboxes as well and where we can hopefully structure Uh How how to improve sandboxes and How to evaluate Them as well so that we can build Uh on on th The Good practices of What has already been established and also to prepare for What's coming next in terms of challenges but More effort more resources need to be done to to put been putting there Uh I invite you to to join Our Uh Our events You Can you can Uh share it You Can you can see
it in Our Website Uh d.org in the part of event and Uh and he see more information around it You're invited I think It's Wonderful to be able to be talking to this particular audience around it we Need More Data protection experts and Data and Ai experts to be around the room but I do think that much more needs to be done so that we actually make the most of the potential Thank you laurine I'll pass Now to Lucia I'll be mindful Time retomo en una lección que al tiempo es un riesgo en el informe
que hicimos En derechos digitales vimos la experiencia de Francia Noruega España que está desplegando su sandbox aplicando la ley de Inteligencia artificial en Europa así también como vimos la experiencia de Francia Noruega no sé si ya los mencioné Qué tienen en común todas estas estas experiencias en Europa y es la ausencia de análisis sobre el impacto de los sandboxes en la protección de los Derechos Humanos de las personas porque Los sandboxes prometen a los innovadores acceder a un terreno seguro de vigilancia no de supervisión de la autoridad pero también de acceso a usuarios de la
vida real usuarios que con seguridad no saben que con ellos se está experimentando con una tecnología cuya seguridad no riesgos todavía no han sido probados o de ellos no se ha generado evidencia luego es un riesgo porque en América Latina cuando vemos estos Desarrollos para avanzar hacia la implementación de sandboxes y experiencias ya pasadas como la de Colombia vemos que los análisis desde la perspectiva de Derechos Humanos también están ausentes luego urge no tener digamos esta perspectiva también sobre la mesa Porque si no vamos sin derechos humanos yo creo que estamos hablando de una promesa
vacía eh particularmente de nuevo porque lo que se busca el innovar En un espacio de juego seguro es permitir a innovadores tener no los datos de personas de carne y hueso reales eh luego cuando las cosas salgan mal Qué haremos con esos usuarios no que en en principio no saben que están siendo parte de un entorno de prueba solo quisiera traer para ustedes una cita muy muy concreta de un autora una solo una autora entre más de 100 textos revisados para la elaboración de Este artículo en derechos digitales hablaba de la importancia de considerar los
Derechos Humanos al hablar de sandboxes y en inglés decía quote unquote regulatory sandboxes generated a lot of hypee but but effect is limited due to the of tested products Services business models furthermore the strong Human Rights guarantees built into process hinder their experimental nature and decrease their attractiveness which is Primarily based on the lifting of certain legal restrictions during the participation in the sandboxes derechos humanos para la promesa digamos de los sandboxes puede ser una limitación y si eso es así vamos a tener que hablar muy en serio de si esta es una herramienta regulatoria
digamos atractiva y en caso de que lo sea Cuáles son las valoraciones de necesidad razonabilidad legalidad en el medio than Seven so I Guess I'll Open for some comments not questions Because We won't have Time for reactions here um but please Go ahead you and theno do you see um the technological competition and regul regulat Race could be obstacle to collaboration in Sand between countries Thank you We Will take notes to follow up after the panel Thank you okay So well very quickly but I couldn't not speak When Uh the brazilian Case was discussed here
so I have been Part on the I have been leading the Team That's working on the design Uh we Still in the design phase of Our sandbox so It's very interesting and Looking Forward to the the report That You're Going To publish Because I agree with you Lucia Uh We are Being cautious and That's actually Connect With Your Last Uh intervention when you mention Uh We have to be careful of Human Rights and especially Data protection Authority Then We have to be Double cautious Because We are a Human Rights a fundamental Rights institution Uh sometimes
other regulatory agencies don't have then officially in their agenda Of course everyone understand Now that Every regulator should care about Human Rights but for Data protection Authority This is in Our It's Our role so This is something to be careful but this does not mean my opinion that Uh we should avoid experimenting Because What I keep saying and playing I told that to All the colleagues We are Still Sand boxing Sand boxes and part of initiatives like the global sandbox Forum I think It's about that let's let's have different levels and experience of sandboxes Because
we can grow maturity and the singapore example is It's excellent example I think the UK as well Now We have a digital coordination Forum like cc and this only happen Because they all the to Experiment so their sandes have sandbox several Times and Now they have a big maturity Now we can try to have coordination Because lacking Now we don't need to be Calling anyone to come Because the participants Trust Us and they want to come so I agree with you Maybe lating America Still Will have some steps to go but It's It's interesting to
See That initiatives are coming and hopeful that maybe with in Five 10 years I don't know but We Will Have Very nice examples Maybe mercosur partnerships on sandbox or Just independent collaboration So let's hope for the best since I literally have the mic like to reinforce the the invitation for the global and Box Forum there's a great panel um on on on the afternoon exactly on that excellent thank so much phip I Guess I started the conversation talking about the topic or the title of this panel Which puts Human Rights or Innovation I Guess the
Take I would like to to to take from this conversation is that we cannot afford to sustain innovation and Human Rights as a dichotomy in latin America there was an element here that was not brought explicitly to the table but Which has to do with inequality precisely and we Are Talk about global inequalities that We are trying to address And I Guess this space has been putting a lot of Focus on that on the The the g20 and the digital sovy and other sovy agendas um and We also have the local inequalities Because latin America
remains One of the most or the most unequal region in the world and When Jonathan talks about aset We Are Not talking only about informational asymmetry we're Talking About different levels of Power asymmetries that affects Our region so I really like I would really like to end this conversation With this reflection on sandboxing Sand boxes Which I really like and the idea that came out of several of your interventions in terms of how do we What is the latin American Take On sandboxes and this Uh I Guess Jonathan you mentioned cc mentioned this in several
Times How do we use this instrument as an instrument for collaboration and as an instrument for building Trust also as Line was mentioning I Guess this was a very valuable panel at least for me as Civil society organization We Are Not often invited to several of this organ of these conversation let's say but We are putting ourselves into the table And I Guess this can be a great Start Uh um of discussion for Us to continue putting the different perspectives into dialogue Because I Guess we can get into very interesting insight Which We are already
hearing here and there in this in this room Thank you so much Thank you especially To lorraine Data fe initiative for this uh uh Last minute merge Which I Guess worked very well Thank you Jonathan Philipp Lucia also for All the work put into building this panel And I Guess we continue in the rooms Ah Thank you so much cc sorry for joining online and Giving such valuable insight into this conversation I Guess We Will continue in the room and hopefully continue to build other spaces to take this conversation on y