Office of the Secretary-General's Envoy on Technology - Press Conference: 27 August 2024
/
36:57
/
MP4
/
2.3 GB
Transcripts
Teleprompter
Download

Press Conferences

Office of the Secretary-General's Envoy on Technology - Press Conference: 27 August 2024


Briefing on global AI governance recommendations by the UN SG's High-level Advisory Body
 
 

Speakers:  

  • Carme Artigas, Co-Chair of the HLAB on AI and former Secretary of State for Digitalization and Artificial Intelligence, of Spain
  • Latifa Al-Abdulkarim, Member of the Shura Council of the Saudi Parliament
  • Seydina Moussa Ndiaye, Lecturer at Cheikh Hamidou Kane Digital University in Senegal
 
Teleprompter
All right, good morning.
I'm so sorry everyone for the brief delay to get started with this press briefing.
[Other language spoken]
I'm part of the Office of the UN Secretary General's Envoy for Technology based here in Geneva.
And I'm delighted to present our esteemed guest this morning for a 45 minute or 40 minute briefing on the process and the early conclusions of the Secretary General's **** level advisory body on artificial intelligence.
I'm joined here by Carmen Artigas, who is Co chair of the advisory body and formerly Secretary of State for Digitalization of Spain.
And to her right is Sedina Moussa, Dr Sedina Moussa, who is a lecturer at Sheikh Hamidou Khan Digital University in Senegal.
And we have perhaps 20 minutes of remarks from our guests followed by your questions.
No, is that correct?
Shorter.
Well, we'll see how much time we have.
And again, apologies for the late start over to Carmel.
Thank you very much, Isabel, for introduction.
[Other language spoken]
We are delighted to be here with you today.
I'm joined here in person by Sedina Musa Nyai.
And it is a pleasure for us to be back to Geneva, this beautiful city which we last saw during winter.
That has contributed in a great deal to our thinking as we had a second here in person meeting of the AI advisory body.
We had our second in person meeting here in March where we met and we consulted especially with many relevant UN agencies who have their own specialised knowledge and expertise in the Nexus of artificial intelligence and development, human rights, health, intellectual property, technology standards, peace, security and meteorology.
That's the the type of UN agencies we met in person last March and a lot has progressed since.
And we're pleased to briefly share with you the progress of the advisory body and our plans for the final report, which is scheduled to be released in mid-september and answer any questions you may have on this process.
Those who wish to have access to the report in advance Ander embargo, please give your contact information to our colleagues from the secretariat to Isabel and she will share and we'll be able to share the final report sometime next week.
So the idea is that if you want to have access to the report in advance and Ander embargo, you should ask Isabel.
So now I will hand over my fellow member Cedina to introduce the work of the body and its progress up to date.
Thank you so much Cedina.
Floor is yours.
Thank you, Carmen.
Hi everyone.
The **** level advisory body is comprised of 39 members across all rail run sectors, 33 nationalities from all regions and it is a gender balance.
We were selected from over 1.8000 nominations including self nominations.
We are multi stakeholder, multidisciplinary and diverse which gives us a unique perspective to address stations of international AI governance.
The body has gone through a rigorous timeline over the past nine months.
We listen to stakeholders at numerous AI events.
We met three times in person in New York, Geneva and and Singapore.
Here in Geneva in March, we visited many relevant UN agencies and diplomatic missions released.
We released an interim report in December, last December and the final report to be released for the public in mid-september.
Our interim report generated much feedback from a course or regions.
We got two 250 plus write on input from 150 plus organisation and more than 100 individuals.
We did a risk pulse check horizon scan of 350 expert.
[Other language spoken]
We did more than we analysed more than 100 papers and did more than 50 consultations and all of this which have been incorporated into our final report as the report will be launched on 19 September.
As our question Carmen mentioned, if you would like to have early access to the report and the embargo, please share your contacts with our colleagues.
Isabel Carmen, back to you.
[Other language spoken]
I think the importance of this report comes from the awareness that there have been a lot of international AI governance efforts, but have really increased since the launch of ChatGPT last October 22nd.
So what we have seen is that while there were like a lot of international efforts, but three initiatives per year, for example, before just in 2023, we have seen 6IN global initiatives around AI governments and just the first half of 2024, we have found 7.
So there's a lot of awareness and interest on different AI international initiatives.
What would we have witnessed and we have assessed during our work in the in the H lab is that even though there are a lot of initiatives today, not neither of these initiatives is globally inclusive.
So existive existing initiatives tend to support the status quo which excludes many states from the decision making process, especially those states in the Global South.
So this correlates with the feedback we received throughout various deep dives.
We have got into depth into 18 deep dives at the approach since impact of AI in children education, in Peas in health and we have had consultations across different regions.
Online submissions confirm that the global governance deficit described in our interim report.
So there is a government, a governance deficit on AI that requires that we have a globally inclusive approach.
So we see several advantages on addressing AI governance issues at the global level.
And basically I would summarise them into three equity, effectiveness and efficiency.
In terms of equity, we need the whole of humanity inclusive governments that we can approach the implications and demands at the global level.
So there is a need for a new social contract in order to share both the benefits and burdens in line with human rights.
And we want to have a side type of government that enables local experimentation, adaptation, autonomy, ownership and that everybody can harness AI opportunities, which are huge.
Then in terms of effectiveness, there is a lack of transparency in information flows, also in knowledge pooling and learning because that is where we need to include also non like minded countries.
We need to enable soft coordination between plurilateral and regional initiatives and to advert and safe race dynamics such as the race to bottom on safety, scope of use, etcetera, as well as we need to empower cross-border interoperability in a governance and standards.
In terms of efficiency, we need a global approach of AI governance to be able to network existing organisation efforts and drowning on existing and pre-existing commitments.
So in our final report we make not one but seven recommendations based on these observations.
Through these recommendations we suggest that the UN complements existing pre lateral initiatives and offers stable and reliable inclusive home for issues that cannot be handled just only among a subset of countries, a subset of companies or a specific domains of UN organisations.
So based on this preliminary conclusions, I would just open the floor to questions that would be moderated by Isabel.
[Other language spoken]
[Other language spoken]
[Other language spoken]
Just a quick correction, we have until 10/25 for Q&A.
If you'd like to ask a question, you can use the chat or raise your hand.
I don't see any hands raised at the moment, I think, but perhaps here in the room we'll have someone break the ice.
Is are you from Associated Press?
[Other language spoken]
[Other language spoken]
Hi, I'm Jamie from Associated Press.
So you mentioned the seven AI, the the you said it take the other one too.
Is this one working?
It should be read, no.
Maybe you can ask your question and I'll repeat it exactly.
I can repeat the question just until we fix the, I'll try to be brief and that'll make it easier for you comprehensive.
What ultimately can the United Nations do to ensure that there is some sort of global governance that standards that countries will actually have to abide by, not simply have it as sort of a, a guidepost?
And, and, and I guess another question is about, about fears about AI.
There are a lot of workers in particular who are worried about the impact of AI.
Are you going to address those issues in your final?
Let me repeat those two questions, Sorry.
[Other language spoken]
One for every single question.
Or we have a let's take a few, maybe let's see how how everybody's feeling.
It's the entree here in Geneva.
It's their own trace.
So everyone's getting back into the swing of things and the technology, but we had two very good questions to kick us off from Jamie of The Associated Press.
So first is, if there is a global governance deficit gap on AI, what could EU NS value add be?
And what is the scope for the UN to ensure that there is enforceability of some of the principles and guidelines that already exist?
And a second question, which touches on fears and anxieties around AI at the moment in the workplace for creators and workers, for creators and workers.
I don't see other hands at the moment.
Oh yes, it we can, we can start these and let's start with that.
So first of all, thank you.
Yes, so I'm going to answer the first question.
[Other language spoken]
I mean, you are really touching the very interesting point.
As you know, in the interim report, we we approach the reason why do we need to global governance and what should be governance in the final report we're going to say how.
So really the final conclusions cannot be shared with before the before the final report is issued.
But it is important to say why do we think UN is the right place to a host of these initiatives And our conclusion is that the UN is the only organisation that has a really proven track record on addressing issues on global governance in many other aspects.
And we see that is what we needed these days.
There are a lot of initiatives.
The AI topic is very complex, is horizontal, it affects all sector.
Most of the tools are have dual use.
There are a lot of issues around safety, but also human rights.
And This is why we think there is important global deficit that has not been addressed and it's not being addressed by the existing international initiatives, not even by the excellent goal that work that the UN agencies do, which is very specific on certain domains.
But because this is horizontal, this effort should be coordinated and we are proposing more holistic approach.
And I think that's where we see the UN having this umbrella role to ensure that that there is efficient in the allocation of resources that is coordination and not overlaps.
And the we and the we first, the first time gathered together private sector, academia, researchers and in governments in such a complex topic.
So I think that not even a single country, not even a single company, not even a single organisation can do it all.
This is why we are proposing a more holistic and collaborative approach.
And I think the UN is the right place to to to set it, especially because under the UN umbrella, we have a very important principle that is also reflected in the interim report, which is everything we do on AI should be under international law, UN Charter and human rights declaration in terms of approaching the fears.
Sorry, do you want to answer this?
[Other language spoken]
About the fields of for the workers.
[Other language spoken]
It's has former technologies.
We will have of course some some job lost.
It's, it's, it's, it's normal, but we will have also a lot of job created.
And I think that it's more an impact of how we will work.
And it's a matter of of adopting the technology so we could improve our way.
[Other language spoken]
So I think that it's more an opportunity than a fee for, for the workers.
Yes, government.
See on this specific topic, as I mentioned, the report will include 18 deep dives.
One of these deep dive includes the impact of AI in, in work.
But of course, you've seen the recent report issued at ILO, Mind the Gap together with the Secretariat and, and I think they are the ones who have the specific knowledge on, on the issues of AI and the potential divides.
I think our recommendation is what are the instruments that we can ensure that these divisions can be solved and where are the places when people need to meet to solve these issues.
So we are not replacing the role that the UN agency is already doing what we're proposing ways in which all the stakeholders that need to might find a solution to those problems need to gather.
And that's what we will see in the report.
But of course, we address these topics.
We have an analysis in every of these topics on the risks, but also on the opportunities, because sometimes recently the discussion is only on the risks.
And I think we cannot forget the great opportunities that are ahead of us.
And I think the importance is how we make that those opportunities are equally beneficial to everybody because we guarantee equally on the access.
[Other language spoken]
And there was a question in the back of the room, if you could introduce yourself, that would be helpful.
Yes, Bobby Siegelson, A freelancer.
So if I may tease you, not quite super, but still accredited.
[Other language spoken]
At the onset of the computer and software revolution in the 80s at Waipu, software was not under patent but under copyright.
It was considered more art and literature than technology.
Not hardware, but software.
Considering that AI is also even more a matter of literature than pure technology, regulating governance of AI sound a bit like regulating literature.
And aren't you a bit disturbed that the governing governance body, the UN, the only one able to either organisation, which has decided to restrict human vocabulary to 17 words?
You're raising a very interesting point around intellectual property that we also had the privilege to discuss with the WIPO organisation here in Geneva.
And they share with us a similar reflection that the barrier between intern industrial property and intellectual property are blurring in the terms of software and AI.
This is also a subject we are approaching in the deep dives in terms of the recommendations.
We are not having a specific regulatory recommendations on the our report because I think that there is also a margin for the nations to to regulate themselves, themselves different aspects of the industry and the market.
But would probably we are aligning claiming is for regulation alignment and that's what we are going to start the the discussions in the recommendations.
How can we make sure that there are not isolated areas of regulation, but there is certain level of regulatory alignment.
That's one of the the aims, the topics we're going to aim with our recommendations.
[Other language spoken]
The floor is open for additional questions.
I looking I see Emma Farge, if you could introduce the the medium that you represent.
Emma, the floor is yours.
[Other language spoken]
Emma Farge from Reuters News.
I'm new to this topic, so sorry if if they're ignorant questions, but were states consulted as part of this process?
And if so, are they generally on board to have new kind of global rules in this area?
Or maybe there are different ones who have different opinions?
If you could kind of talk me through that aspect of this and maybe I missed it, but would these, would this governance be binding in any way?
And and how would countries and companies be be bound by new rules in this area?
Thanks very much.
Let me briefly repeat those for karma.
We forgot to put these things on to, to hear your excellent questions, Emma.
So the first one is, were Member States consulted in this process and are they on board for these ideas?
And the second question is, are the proposals of the advisory body binding over to you?
So our Board is a totally independent body.
We are not representing individually any corporation, any, any country.
I myself used to be the former estate strategy of Spain, but I'm not representing Spain at all.
I'm acting on my personal capacity also Cedina on all the 79 members.
So it's a very independent report which issues recommendations that people can decide to adopt or not.
And parallel, there is of course a political process related with a global digital compact.
And we have shared a lot of information during all this process with member states.
But in the preparation of the report, what we have consultant is experts, independent expert, hundreds of them, I think, I would say even thousands, I think more than 75 meetings on deep dives.
So because of the terrigenity, the diversity in the background, academia, policy makers, technology companies, we have come and that has been a really interesting work and and complexity come with a consensus of the state-of-the-art on the issues and the recommendations.
So you should interpret our our recommendations are absolutely scientific based, evidence based.
There's a very important in depth intellectual analysis and all the recommendations, but independent.
So I think what we have done is to share all these recommendations with the member states to the different information meetings.
And now as we all know, there's there's a parallel, you know, discussions on the global digital compact and we expect that, that there's a positive feedback on that, but we are not aware of that process as as the body itself.
And perhaps for Cedina, do you expect that the recommendations will be binding of the report binding binding.
So are they obligatory or how would they eventually be implemented?
[Other language spoken]
Yes, I think that as as we share all the recommendation with the Member States is for each Member States to, to, to take into account these recommendations.
And I think that there is some, there is some recommendation.
We will be talked at the summit, the global, the, the summit of the future and we, we, we, we, we ask all the Member States to, to be supportive about our recommendations because the recommendations we are providing could help to, to have a global, a global governance with, with some spaces for global S to, to share their voices.
So I think that it will be, yes, it will be great.
I would only include that I mean the feedback we received during this conversation was very positive and don't forget that we are consultancy body for the Secretary General and it is the Secondary General, the one who is taking these considerations, these recommendations into consideration or not and therefore make a formal proposal to the Member States.
[Other language spoken]
We have Isabel Sacco from FA and followed by Maya Plants.
[Other language spoken]
I would like to clarify some organisational issues on for example, this final report will be presented anywhere else next week or New York or somewhere else or this is for you that the like the official launch even if it's under embargo.
Otherwise, it would be it would have been interesting to have the report under embargo to be able to ask questions regard related to the feasibility of your recommendations, you know, and if you can maybe send us through the information service here in in Geneva.
The report on the under embargo to not to be.
I mean to everyone has to then to, to ask everyone by themselves if you can just send to the, to the, to the Geneva press corps the the the report and their embargo.
It would be nice.
[Other language spoken]
Thanks, Isabel for your question.
Yes, that I've already had four emails this morning from the press asking for the report.
So your point is well taken and we'll make sure to facilitate as much information as possible.
And we have a follow up question from Maya Plants and this is from the UN brief.
[Other language spoken]
Yes, thank you for taking my question.
[Other language spoken]
I don't know if you mentioned in the very beginning, but they UNESCO just published the consultation paper on AI regulation, emerging approaches across the world on in AI governance.
And one is principles based, the daughter standards based, agile and experimentalist approach, facilitating and enabling approach, adapting existing laws approach, access to information and transparency mandates approach, risk based approach, rights based approach and liability approach.
[Other language spoken]
The office is working.
Your office is working with UNESCO in this approaches.
[Other language spoken]
And then a second question I have is regarding, don't you think there's too much participation of the technology companies that they are shaping these regulatory approaches, a little too much in the case of the advisory body for the UN Secretary General?
First of all, clarify kind the question for the representative from F, this is not the press release on the report.
This is a previous press conference to update you on the progress of the our work.
As you have mentioned and my colleague has already said, the official report will be issued on September 19th.
As of September 13th, we're going to distribute the final report under embargo.
And yes, it will be an official presentation during the UN, GA, during the summit of the future, probably Sunday 22nd.
We're fine tuning the date, but that day will be a official presentation together with the Secretary General and there are some events during the summer of the future we can to to share publicly the conclusion.
So there will be an important way to officialize the report and you might have access in advance and embargo for this report.
So the, IT will be published on the 19th, but the official presentation will be during the summit of the future and the UNGA week.
Apart from this, you were mentioning my, the, the role of UNESCO.
The, you know, the role of UNESCO has been very, very valuable.
You know, our, our inputs to, to the report as well the role of ITU and all the existing UN agencies.
You know, ANESCO issued the first ethical guidelines, but when we talk about governance, we go beyond ethical is what should be done and governance is to be how do we sure the people are doing the right thing.
We don't see governance as an innovator, but as an enabler of doing the right things first time.
And that's why we have involved the different stakeholders, all these reports.
We're also glad to see that also our work is inspiring the the, the work that UNESCO and the other agencies will do in the future because we need them to do their work in the future in the eyes of AI.
So what we see is there a vertical specialisation, Every agency, I mean, who better than ILO to assess the risk of the impact of AI in the world?
Who better the wiper to assess the impact of AI in IP?
We are not trying to do the work of these agencies 24/7.
We're creating the mechanisms that all these complex problems have, the complex solutions we need, putting the stakeholders together.
So our recommendations are not solving every single question and every single matter.
We are saying who should meet to find the solution.
We're 24/7 to find the solution to this topic and that's what we are going to propose the right mechanisms to do all these efforts aligned and to solve all these issue.
And therefore, of course there are some institutions that have more influence in governments that are UNESCO, but have probably less influence in the private sector.
And we need to be able to to find the mechanisms that everybody is working in the same direction.
I just don't don't share with you the view, at least on my experience that we have an over representation of technology companies in the world at all.
I would say have a very important contestancy of NGOs and academia.
I don't know Cedena, what is your your feedback?
But we see that the pressure or the different views are equally represented.
I'm yes, Co chair that sometimes need to be like the the referee.
[Other language spoken]
But at the end of the day, the trade off is bringing us the best of the both options.
So I'm quite happy with the representation and diversity of the board.
Celina yes, thank you.
Kwame indeed in the in the board.
I think that having technology companies was interesting because when we say we have to to, to bring everyone, I think that bringing the, the, the, the point of view of, of global S, the point of view of member states, the point of view of NGO and academics.
It is also important to have the point of view of technology companies.
And the work in the in the body was yes, was very effective.
And I think that there's no problem with having them inside.
I still see the hands from Isabel, Sacco and Maya up on the zoom.
I don't know if those are additional questions or just the handset, the buttons of zoom that are sometimes hard to find.
In the meantime, I'm over to you.
Oh, sorry, was not excuse me now you I, I don't have any follow up.
[Other language spoken]
[Other language spoken]
Then there's an additional hand up here in the room and we have about four more minutes for the briefing.
Jamie, over to you.
My hand is a real hand, not a virtual one.
The I just wanted to to follow.
I mean, I mentioned before like the fear of jobs and on jobs and creation.
The flip side is, is that in recent months we've seen a sort of a, a retraction or a decline in the amount of concern that I actually is going to do as much damage as it might have been.
So my question is, and you sort of alluded to this about this, but are you not concerned that maybe you're not the, the, the AI world is not clear enough yet.
We still don't know exactly how it's going to shift and shape out and that the fears that we have today may not be there in six months from now.
The the, the, the, the marketplace is moving and the world of AI is changing so fast that it's you maybe even premature by putting a finger on it right now.
We have one more question and then a few minutes for say Tina and Karma to wrap up Maya plans.
[Other language spoken]
I have a follow up following on Emma Farges question regarding countries, how much governments have been shaping this conversation, this advisory body, which countries are supporting of a lesser fair approach and which ones are really more concerned about the negative impacts?
And then the second question is we see a consolidation of several companies have been acquired already by the big tech companies like Entropic Maiden, Acqui Hire with Suleiman being invited to become the chief AI person at Microsoft.
And we have several other because it seems like AI, generative AI hasn't been yet found, has not yet found a viable business model and it costs billions to actually just make people search for a couple of things ruin.
They prompt something like millions of litres of water are being used.
How are you addressing this question of the responsibility that this big tech companies have in terms of addressing the environment question instead of just greenwashing?
OK, thank you for the question.
I, I, I think that you're right when you say that we don't know how we'll be the feature of, of AI.
So it's why it's important to, to talk about the global governance, a way to think all together about the future and how to tackle the, the future problem.
So yes, I just want to add that, I mean the question about risk is a very relevant 1.
And This is why in the final report we are including a specific agenda on what we call the risk pulse.
We have made really important consultation, very broad consultation on what are the perceived risks.
And you will see how interesting is that the perception of risk is different from one countries to the other.
It's even different from between women and men.
And we're talking about the risk we need to have to differentiate short term risk and long term risk.
If we remember the very early days of AI development, everybody was talking about fairness, that was talking about bias, discrimination around algorithms.
Then this conversation shifted one year ago on safety because of the AI safety summits.
And we're like we're thinking about, we're not worried about fairness anymore, we're not worried about.
So this conversation will go on and on and on.
And it's a very dynamic environment.
So in our approach, we take both into account short term, mid term and long term term risks and how do we keep that sustainable depending on the different future technology waves.
And as I mentioned, it will be very relevant to, to know the outcomes on the response because then we see also how also I would say lack of inclusivity is, is perceived also in the risk the population have internally.
And yes, following the conversation with Maya, I agree with you.
There are some hypes.
Now the big thing is generative AI, we'll see a big boom consolidation.
That's how it grows in the technology sector.
I've, I'm, I'm from the talent agency sector for 30 years and I've seen it many, many times.
But the important thing is how can we keep these companies accountable?
And really where the global deficit is, is in accountability and reporting on transparency of the R&D from private companies because the first time ever technology is not developed mostly from the academia, but mostly from R&D centres in private companies that are not obliged to publish their information.
So one of the the topics we approach is how can we show transparency on the opportunities and risk on the continuous way over the time and how we set up monitoring systems that make companies accountable in terms of and not only social responsibility, but also on sustainability, which is also a very important topic that we approach in the document.
So I totally agree with you.
We need transparency, we need accountability.
[Other language spoken]
Thank you so much to all of you who have joined us online and especially to our colleagues at UN Information Service.
It's a pleasure to answer your question.
So Carmen Sedina are here for the rest of the day in case you'd like to chat with them bilaterally.
And you have my e-mail on the chat if you need more.
Everyone has said they need more information, so I'll agree with you and Information Service how we get that to you.
And again, thank you.
Have a wonderful day.