Sardine named a Leader in The Forrester Wave™: Financial Crime Management Solutions, Q3 2026

Learn More
The Saturday Fraud Strategist

Fraude amigável no comércio agêntico: por que cada comprador ganha uma nova desculpa

10 min

Todas as conversas sobre comércio agêntico acabam na mesma pergunta: quando um agente de IA compra a coisa errada, quem paga? Investiguei o assunto há algum tempo e ainda não tenho uma resposta clara. Quando cheguei ao Money 20/20 Vegas, fui falar com alguém que vê, todos os dias, como essa questão se resolve na prática em disputas.

Ben Herut é vice-presidente de Pagamentos na Chargeflow. Desta vez, fomos breves. Falamos sobre quem assume a responsabilidade, o que os emissores e os comerciantes precisam de saber quando um agente faz compras, por que motivo o 3D Secure tem dificuldades sem intervenção humana e como a fraude amigável no comércio agêntico pode dar aos consumidores uma nova desculpa.

O que vai ouvir neste episódio:

  • Quem é responsável pelas compras feitas por agentes de IA e porque é que a Visa, a Mastercard e a AmEx ainda não resolveram esta questão
  • Por que o fornecedor de IA é um novo interveniente na cadeia de pagamentos e como o modelo «compre agora, pague depois» (BNPL) oferece uma comparação útil em termos de responsabilidade Por que os emissores precisam de identificar as transações realizadas por agentes de IA e o que as autorizações de pagamento, de cartão e de intenção podem proporcionar Por que o registo de agentes de IA junto de emissores e comerciantes é importante e o que significa ter uma loja online preparada para estes agentes
  • Porque é que o 3D Secure pressupõe a intervenção humana e o que acontece quando um assistente faz uma compra sem avisar o seu utilizador
  • Como a fraude de primeira parte aumenta quando um comprador pode dizer que o erro foi do agente de IA
  • Por que os incentivos à segurança dos fornecedores de IA são importantes e quem tem maior probabilidade de pagar a conta enquanto eles permanecem pouco claros

Deve ouvir este episódio se:

  • Gere estornos ou contestações num comerciante, adquirente ou emissor e quer preparar-se para o comércio agêntico
  • É responsável pelos fluxos de autenticação ou de 3D Secure e questiona-se sobre como funcionam quando um agente faz compras
  • Querem compreender quem poderá ser responsabilizado quando um agente de IA faz uma compra
  • Estão a planear como comprovar o que um comprador aprovou quando um agente agiu em seu nome
Notas do episódio e principais conclusões

Ainda não se sabe quem é o responsável

Ben diz que a maior questão em aberto é saber quem assume a responsabilidade. Ainda ninguém sabe. A Visa, a Mastercard e a AmEx estão a procurar respostas, embora ele descreva a situação como um passo atrás, porque há agora um elemento imprevisível na cadeia de pagamentos. Acrescentei que o fornecedor de IA é, na prática, um novo interveniente nessa cadeia e perguntei por que razão só o comerciante ou o emissor deveriam assumir o risco. Ben compara a situação ao BNPL (compre agora, pague depois). As empresas de BNPL nunca foram a parte responsável, e a responsabilidade era transferida em função dos contratos assinados. Ele não espera que os fornecedores de IA sejam responsabilizados e, pela sua experiência, também não espera que os emissores o sejam. Tudo depende dos dados que circulam no sistema. Até que se defina a transferência de responsabilidade no comércio com agentes de IA, essa incerteza afeta todos os intervenientes da cadeia.

Os emissores e os comerciantes precisam de ver o agente

No início deste ano, as redes de cartões pareciam hesitar em informar os emissores se uma transação era realizada por um agente de IA. Desde então, recuaram nessa posição, embora não seja claro como ou quando os dados serão partilhados, nem o que os emissores poderão fazer com eles. Ben destaca três mandatos nos pagamentos realizados por agentes de IA: o mandato de pagamento, o mandato do cartão e o mandato de intenção. Sem essas informações, é difícil para qualquer uma das partes tomar uma decisão. Na sua opinião, ambas as partes precisam de ter visibilidade sobre o agente. O emissor deve saber que uma transação é realizada por um agente de IA e ter um pouco mais de cuidado. O comerciante deve poder demonstrar que está a lidar com um agente, e a sua loja online deve ser fácil de analisar pela IA. O agente tem de estar visível para que as regras possam funcionar.

3D Secure e um sistema de resolução de litígios concebido para humanos

O 3D Secure pressupõe que é uma pessoa que está a efetuar a transação. Um assistente pode fazer uma compra sem avisar ninguém, e o titular pode estar no dentista, num avião ou a dormir. Na minha opinião, isso põe o sistema em causa. Ben foi mais longe. Todo o ecossistema de estornos e contestações foi construído com base no comportamento humano, e nada estava preparado para um novo interveniente. Aponta a transição da PSD2 para a PSD3 como um exemplo de como a regulamentação tenta acompanhar a evolução e afirma que está sempre atrasada. Por enquanto, não existe uma solução única de autenticação para transações realizadas por agentes de IA. O cenário muda quase todas as semanas. Devemos contar com alterações às regras durante algum tempo.

A fraude amigável ganha uma nova desculpa

É na fraude cometida pelo próprio comprador que os estornos no comércio por agentes de IA se tornam um assunto delicado. Ben diz que cerca de 70% dos estornos são casos de fraude amigável ou contestações inválidas, divididos entre casos em que o comprador realmente não reconhece a compra e contestações intencionais. Agora acrescente um comprador que diz que esperava ter dado a instrução certa ao agente. Brinquei que essa é a desculpa perfeita para o arrependimento após a compra: foi uma alucinação da IA. Certa vez, dei o exemplo de pedir a um agente que comprasse um iPhone e vê-lo comprar uma foto de um iPhone. Ben diz que as compras por agentes ainda estão no início e que, em algum momento, todos terão de lidar com contestações de compras feitas por agentes de IA. Também observei que a Amazon não permite que assistentes de IA façam compras. A resposta de Ben foi que talvez eles saibam algo que nós não sabemos.

Quem paga é quem investe

Se os fornecedores de IA não tiverem motivos para impedir a apropriação de contas, o roubo de identidades ou o roubo de credenciais de API, teremos introduzido um interveniente cujo principal objetivo é ganhar dinheiro. Não vejo como isso possa acabar bem com o modelo atual. As redes de cartões estão a recuar para observar, e alguém vai pagar o preço. O Ben concorda que provavelmente serão os comerciantes. Até que os incentivos estejam alinhados, todas as equipas devem preparar-se para essa possibilidade.

Conclusão final

A fraude amigável no comércio com agentes de IA não cria propriamente novas formas de fraude, mas muda quem pode ser responsabilizado. A fraude amigável já representa a maioria dos estornos, e um agente de IA dá ao comprador uma desculpa muito fácil de usar. O sistema que resolve essas contestações foi concebido para o comportamento humano. A responsabilidade, a autenticação e a partilha de dados continuam a ser questões em aberto. As equipas que se prepararem com antecedência não vão esperar que as redes de cartões deem essas respostas. Vão definir como comprovar o que o comprador autorizou e o que o agente tinha permissão para fazer.

Recursos e ligações

Ainda não quer terminar a conversa sobre o meu tema favorito, que espero que também seja o seu? Subscreva a newsletter O Estratega de Fraude de Sábado.

Conecte-se com Ben Herut | LinkedIn
Chefe de gabinete na Chargeflow
Colunista no FraudBeat
Membro do Comitê de Fraudes do Merchant Risk Council
SQL, Reg E, 3D Secure, Reclamações e Resolução de Chargebacks

Conecte-se com Chen Zamir | LinkedIn
Apresentador do The Saturday Fraud Strategist
Ajudando fintechs a desenvolver defesas mais inteligentes contra fraudes
Coautor de “O guia de IA para quem combate fraudes”

Episode transcript
Chen Zamir
Chen Zamir
00:06
Hi everybody and welcome to a special episode
Chen Zamir
Chen Zamir
00:09
of The Saturday Fraud Strategist. Here live from Money 2020 Vegas.
Chen Zamir
Chen Zamir
00:13
With me I have Ben Herut VP of payments at Chargeflow. Welcome Ben.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
00:18
Thank you so much. Thank you for having me.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
00:19
It's always a pleasure to be there with all colleagues.
Chen Zamir
Chen Zamir
00:22
Yeah. Uh we have a special bearded episode uh today.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
00:26
Absolutely.
Chen Zamir
Chen Zamir
00:27
Yeah. Uh Ben, so I wanted to speak to you about agentic commerce.
Chen Zamir
Chen Zamir
00:31
Um so I you know I spent some time looking into this.
Chen Zamir
Chen Zamir
00:34
I wrote even a a paper about it uh I think
Chen Zamir
Chen Zamir
00:37
about like 6 or 12 months ago.
Chen Zamir
Chen Zamir
00:39
But you're one of the experts in the industry.
Chen Zamir
Chen Zamir
00:42
Uh so I wanted to grab you and ask you a couple of questions.
Chen Zamir
Chen Zamir
00:46
And you know like let's start with concerns.
Chen Zamir
Chen Zamir
00:49
Uh there are a lot of open questions about uh agentic
Chen Zamir
Chen Zamir
00:53
commerce at the moment, uh and a lot of unknowns.
Chen Zamir
Chen Zamir
00:56
And I'm just wondering you know when you're looking at the space
Chen Zamir
Chen Zamir
01:00
right now what are the biggest question marks that you are concerned
Chen Zamir
Chen Zamir
01:05
about or worried about that are not answered yet?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:14
Well I think the biggest one is who's liable I think nobody actually
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:18
knows I think Visa and Mastercard and AMX are trying to get there, right?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:22
But as always, it's a step back.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:24
And at the end of the day you're throwing
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:26
into the equation a wild card, right? A joker.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:29
And nobody actually knows how it's going to perform.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:32
Obviously, there are testing and so on, There are specific partners.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:36
But at the end of the day, someone is doing actions that were
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:40
made by humans in the past, and now we need to adapt.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:43
And the whole ecosystem needs to understand
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:45
that now there's a new player in town.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
01:48
And how do we how do we define who who's at fault, who's liable?
Chen Zamir
Chen Zamir
01:52
No, it's a great one.
Chen Zamir
Chen Zamir
01:53
Because I think it's also like we're not officially, but sort
Chen Zamir
Chen Zamir
01:57
of, introducing a new actor into the payment chain, right?
Chen Zamir
Chen Zamir
02:01
Which is the actual AI provider, right?
Chen Zamir
Chen Zamir
02:03
Um, so are they liable at certain cases?
Chen Zamir
Chen Zamir
02:06
Why is it just the merchants or the issuers?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:11
Absolutely.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:11
And usually when I talk about agendic commerce, I compare it to BNPL.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:15
Because it had the same type of influence on on the industry.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:19
Because at the end of the day, you've got the BNPL
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:21
companies themselves, but they were never liable.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:24
Was it It was still in between the issuer or the merchant, right?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:28
The liability shifted based on the contract that they were signed.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:31
But here again, you've got the agentic commerce.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:34
Are OpenAI going to be liable for something?
Chen Zamir
Chen Zamir
02:37
Exactly.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:38
Probably not.
Chen Zamir
Chen Zamir
02:38
Exactly.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:39
Is the issuer going to be liable?
Chen Zamir
Chen Zamir
02:40
Yeah.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:40
Mmm, from my experience probably not.
Chen Zamir
Chen Zamir
02:42
Yeah
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:43
But it depends.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:43
I guess it depends on the type of uh information
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:46
that run through the system.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
02:47
But again, the biggest issue is that nobody actually knows.
Chen Zamir
Chen Zamir
02:51
And and I saw that, like to a degree, it seemed like earlier in
Chen Zamir
Chen Zamir
02:55
the year that the schemes are apprehensive around uh sharing
Chen Zamir
Chen Zamir
02:59
with issuers um whether a transaction is agentic or not.
Chen Zamir
Chen Zamir
03:02
But it seems that they walked it back to a degree and now they are.
Chen Zamir
Chen Zamir
03:07
Uh but it's not really clear how and when and um
Chen Zamir
Chen Zamir
03:10
and what issuers would be able to do with it.
Chen Zamir
Chen Zamir
03:13
Like how do you see it at the moment?
Chen Zamir
Chen Zamir
03:16
What what can issuers when they look at agentic commerce and you know
Chen Zamir
Chen Zamir
03:20
that that's something that for them is very hard to deal with, right?
Chen Zamir
Chen Zamir
03:25
Uh you know they haven't been or they are not in
Chen Zamir
Chen Zamir
03:28
direct uh uh contact with the the shopper.
Chen Zamir
Chen Zamir
03:31
Um and even with 3D secure um when know 3D secure you
Chen Zamir
Chen Zamir
03:34
can relay the device, uh fingerprint, and so on.
Chen Zamir
Chen Zamir
03:38
Even that is now no longer in the game because you have agentics.
Chen Zamir
Chen Zamir
03:42
How do you like how would issuers uh play?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
03:58
Right.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
03:58
So again at the end of the day uh when we started talking about AI uh
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:03
payments right that they usually spoke about three different mandates.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:08
You got the payment mandates, the card mandates, and uh you know
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:12
the basically the approval itself or the intention mandate.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:16
Um as long as you are not sharing this info it's going to be very
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:20
difficult for anyone to do, to make a decision on top of it.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:24
At the end of the day if I own agents right and I I operate them.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:28
I allow them to buy something on my behalf.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:31
It has to be not only clear what I'm allowing them to
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:34
do, but also it has to be registered on both sides.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:38
So the issuer knows this is an agentic commerce.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:41
I need to be slightly more careful, potentially.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:44
And also once it's hit the merchants, right?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:47
They need to be able to show that they are
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:50
interacting with a with a with an agent.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:52
So they have to make sure that their storefronts are available for
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
04:56
AI to easily skim through and find whatever they need to find.
Chen Zamir
Chen Zamir
05:01
So okay, let let me let me uh pick up on this.
Chen Zamir
Chen Zamir
05:04
So as an issuer, you know, one of the main thing I
Chen Zamir
Chen Zamir
05:07
I mentioned it uh a second ago, one of the main tools
Chen Zamir
Chen Zamir
05:10
that you have to control fraud is 3D secure.
Chen Zamir
Chen Zamir
05:13
Um but 3D secure you know assumes that you have a human in the
Chen Zamir
Chen Zamir
05:17
loop meaning a human uh uh uh placing the transaction.
Chen Zamir
Chen Zamir
05:21
When an agent does that it might be that
Chen Zamir
Chen Zamir
05:24
the human is is not even in the loop.
Chen Zamir
Chen Zamir
05:26
Now now we have uh these these Muse and Dots and all of these
Chen Zamir
Chen Zamir
05:30
kind of like assistants AI assistants that won't necessarily
Chen Zamir
Chen Zamir
05:34
even alert you before they base a transaction.
Chen Zamir
Chen Zamir
05:37
So, a human might not be, you know, like
Chen Zamir
Chen Zamir
05:40
aware of their phone or whatever.
Chen Zamir
Chen Zamir
05:42
I always give the example, maybe they're
Chen Zamir
Chen Zamir
05:45
in at the dentist at the moment.
Chen Zamir
Chen Zamir
05:47
Maybe they're on a plane, maybe they're asleep.
Chen Zamir
Chen Zamir
05:50
How like how does 3D secure uh come into play here?
Chen Zamir
Chen Zamir
05:54
Like in my mind, it's it's you know, it's broken.
Chen Zamir
Chen Zamir
05:57
Uh how how do you how do you uh introduce authentication
Chen Zamir
Chen Zamir
06:01
flows in agentic uh commerce flows?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:05
So that's a very tough question to be honest.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:07
And I I'll take it even a step further because the whole chargeback
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:11
or dispute ecosystem was built on human behavior.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:14
None of it was ready to to to have this new actor in it.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:18
Um 3D secure is one one of the aspects.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:20
But as you can see for example we've got the PSD2 that transferred
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:24
into the PSD3 and they made the relevant changes into it.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:28
So it's always lagging.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:29
But at the end of the day they would find a way to make sure that the
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:33
the new data, or the new performance of how each one of the actions is
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:38
taken on the ecosystem, would be monitored and taken into consideration.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:42
As far as I know, there's no there's no one answer currently.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:46
But it's it's still believe and I think it's very important to
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:50
look into and hopefully we're going to have more info soon.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
06:53
It's changing on a weekly basis almost.
Chen Zamir
Chen Zamir
06:56
Yeah, that's super interesting.
Chen Zamir
Chen Zamir
06:58
Let me flip it now and ask about first party fraud.
Chen Zamir
Chen Zamir
07:01
Because I I think with third party fraud, I mean
Chen Zamir
Chen Zamir
07:04
I understand the the the issues or like the how
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:08
Complexity
Chen Zamir
Chen Zamir
07:09
Yeah, the complexity. And how the chain breaks.
Chen Zamir
Chen Zamir
07:12
But with first party fraud, it's you know it's it's a bit bizarre.
Chen Zamir
Chen Zamir
07:16
We we are very focused on AI hallucinations and whether they can, you
Chen Zamir
Chen Zamir
07:20
know I I you know once gave the example you know buy me an iPhone and
Chen Zamir
Chen Zamir
07:25
they buy a picture of an iPhone right as an as an as an hallucination.
Chen Zamir
Chen Zamir
07:29
But what happens like what are the different scenarios that
Chen Zamir
Chen Zamir
07:33
can happen with first party fraud and agentic commerce.
Chen Zamir
Chen Zamir
07:37
And how should merchants or issuers uh think
Chen Zamir
Chen Zamir
07:40
about it as as they try to prepare for it?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:43
So I I don't want to have anyone too scared
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:45
but basically it opens a Pandora box.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:48
Because until now we I mean the the the basic statistic is
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:52
70% of chargebacks are uh friendly fraud or invalid.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:55
And these are usually in in two different sections.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
07:59
One of them is is uh unknown real unknown uh uh dispute
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:02
reason, and the other one is intentional. Right?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:05
So now when you add the layer of complexity of the someone else
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:10
interacting on my behalf, and I hope I made the correct prompt
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:14
to the agent to buy the correct stuff that I wanted.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:17
Now you just a roomful uh basically denying
Chen Zamir
Chen Zamir
08:21
Yeah
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:22
This was me. Is just that much bigger.
Chen Zamir
Chen Zamir
08:25
It's the it's the earth excuse for buy remorse. Right.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:28
Absolutely.
Chen Zamir
Chen Zamir
08:29
Yeah. I didn't mean that. It was, you know, it was a hallucination.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:32
Absolutely. So, it just opens up a new world of denying, right?
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:36
Um, and this is a challenge that everybody is going
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:39
to encounter at a certain time in place.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:42
I mean we know that nowadays agentic permits are just in
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:46
the you know starting, and it's not that wide common.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
08:49
But it's going to hit everyone.
Chen Zamir
Chen Zamir
08:53
Which is interesting because we're seeing that,
Chen Zamir
Chen Zamir
08:56
you know let's take Amazon as an example. And Amazon is a trend setter.
Chen Zamir
Chen Zamir
09:00
They don't allow AI assistance To shop
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
09:03
Maybe they know something we don't know
Chen Zamir
Chen Zamir
09:04
Yeah yeah
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
09:06
But how do you guys look at the agent payments?
Chen Zamir
Chen Zamir
09:10
Agentic payments? Well I said completely the same.
Chen Zamir
Chen Zamir
09:13
Uh yeah, I mean there's the KYA aspect and yeah, there's like
Chen Zamir
Chen Zamir
09:17
the body detection aspect, but I think in my mind the the point
Chen Zamir
Chen Zamir
09:21
that you made about liability shift and where it lies.
Chen Zamir
Chen Zamir
09:25
And you know, liability in the end, it's you know who pays
Chen Zamir
Chen Zamir
09:29
is also who would invest in the solution, right?
Chen Zamir
Chen Zamir
09:32
So while we don't have that, and while by the way while
Chen Zamir
Chen Zamir
09:35
we don't put any motivation on the AI uh um labs or the
Chen Zamir
Chen Zamir
09:39
AI providers themselves, to invest in security.
Chen Zamir
Chen Zamir
09:42
In uh being able to you know like block uh ADO, block uh
Chen Zamir
Chen Zamir
09:46
stolen identities, block um stolen API credentials.
Chen Zamir
Chen Zamir
09:49
While we don't give them the motivation, We are introducing
Chen Zamir
Chen Zamir
09:53
a player that just wants to make money.
Chen Zamir
Chen Zamir
09:56
Uh and I don't I don't see it, I don't want to say I don't
Chen Zamir
Chen Zamir
10:00
I don't want to sound an apocalyptic as you said.
Chen Zamir
Chen Zamir
10:03
But I don't see it ending well with um with a setup right now.
Chen Zamir
Chen Zamir
10:07
And I agree with you that you know the schemes kind of
Chen Zamir
Chen Zamir
10:11
like are taking a step back to watch how things play.
Chen Zamir
Chen Zamir
10:14
And someone would pay the price for it,
Chen Zamir
Chen Zamir
10:17
which is probably the merchants.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
10:33
Unfortunately. Yes,
Chen Zamir
Chen Zamir
10:44
Ben, it was a pleasure having you.
Chen Zamir
Chen Zamir
10:46
We should uh do something like proper on the pod itself.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
10:39
Oh, of course. It was a pleasure. Thank you so much for inviting me.
A man with a long, dark beard wearing a black Chargeflow t-shirt against a dark background.
Ben Herut
10:42
Always a pleasure. And yeah, let's see what happens.
Chen Zamir
Chen Zamir
10:45
All right. Thank you very much. And folks, I see you next Saturday.