0:03

The lack of transparency is a

0:05

rife area of risk for biased

0:07

responses. The whole discussion about AI

0:09

ethics is whether you see it as

0:11

a value or whether you see it

0:13

as compliance. My comparison I use is

0:15

with health and safety. Whatever values

0:17

you have in the analog world should be

0:19

extended to the digital world. Hello

0:25

and welcome to Insight Story, tech trends

0:27

unpacked for business leaders. The podcast that

0:29

gives you the insight you need to

0:31

make the right decisions about how to

0:33

use tech to benefit your organisation. I'm

0:36

Susie O'Neill and this is brought to

0:38

you by Kaspersky, the cyber security specialists.

0:41

This time we're asking, what

0:45

do you need to know to use AI ethically?

0:52

Now we live in a world where algorithms

0:54

can make decisions and data fuels innovation. It

0:57

means that thinking about ethics is more critical than

0:59

ever, especially for anyone in business.

1:03

Organisations need to balance the benefits of

1:05

this incredible new technology with making sure

1:07

they preserve their own integrity and

1:09

protecting their customers. How

1:11

does possible AI bias affect businesses?

1:14

What can you do to ensure fairness? And

1:17

what about privacy and the responsible use of

1:19

data in this pursuit of progress? To

1:22

help me answer these and other questions, I'm

1:24

going to be joined by two people who work at

1:26

the centre of these issues. Tomoko

1:29

Yokoi is a researcher based at

1:31

the IMD Business School in Switzerland

1:33

and she writes extensively about digital

1:35

transformation. And Andy Crouch

1:37

is a tech entrepreneur and consultant for

1:40

Acumen, a world leader in ethical use

1:42

of audience data. But

1:44

of course, we need to make sure we're all on the same

1:46

page when it comes to all of this. And

1:49

we have some very ethical help.

1:51

Hello, I am Dolly, your insight

1:53

assistant. Dolly, thank you for coming

1:55

by. Now there's a lot of

1:57

noise about the ethical side of artificial intelligence and

1:59

business. What's the main worry? The

2:02

key concern is around fairness and fears

2:04

of bias in AI algorithms. This

2:07

can affect decision making and deepen

2:09

societal inequalities. Right, I

2:11

can see that's an issue. What kind

2:13

of problems can this lead to? As

2:15

an example, facial recognition systems have

2:18

shown biases against certain groups of

2:20

people which could impact their access

2:22

to services. Biases in

2:24

job hiring algorithms can disadvantage

2:26

some people. Amazon's recruiting tool

2:29

was found to favor male applicants. And

2:31

what other concerns are there? Inaccurate

2:33

data can lead to hallucinations in

2:35

large language models in generative AI

2:38

technologies that power chatbots like

2:41

ChatGPT or GoogleBard. There

2:43

are transparency concerns around the data

2:45

sources used in Gen AI, also

2:47

how much human oversight is needed

2:50

in AI-driven systems and whether companies

2:52

prioritize automation over human employment.

2:55

Yeah, we need to keep humans in the loop, wouldn't you

2:57

agree? Anything else? It's vital

3:00

that businesses protect users' privacy and

3:02

make sure customer data is secured

3:04

and not accessed without consent.

3:07

When interacting with customers, organizations

3:09

need to be clear about whether

3:11

people are talking to a human or to

3:13

a bot. Yeah, I'm beginning to see what

3:15

all the fuss is about. Are companies

3:17

already using this technology? The

3:20

latest McKinsey Global Survey on the current

3:22

state of AI says one-third of respondents

3:24

are using Gen AI regularly in at

3:27

least one business function. It

3:29

is worth noting that it is only

3:31

one year since generative AI tools became

3:33

widely available. Yay, thank you.

3:36

So, Dolly, should we label you to say

3:38

whether you're a human or a bot or

3:41

are you a prototype of artificial general intelligence?

3:43

Goodbye. Oh, okay, touch to the

3:45

nerf again. Unfortunately, AI doesn't

3:47

have all of the knowledge yet, so

3:50

clearly we need to find some answers. Fortunately,

3:53

I'm joined by two people with

3:55

insight into exactly these kind of

3:57

issues. Firstly, Tomoko Yukoi is a

3:59

senior business executive and now a

4:01

researcher based at the Global Centre

4:03

for Digital Business Transformation at the

4:05

IMD Business School in Switzerland. She

4:08

writes extensively on issues surrounding

4:10

digital transformation including AI and

4:12

ethics in the corporate world.

4:15

Hello Tomoko. Hello. Anani Crouch

4:17

is a two-time tech co-founder

4:19

and now a business development

4:21

consultant including for Acumen, a

4:23

world leader in textual data

4:25

analysis using software to identify

4:27

emotions, behavioral drivers and topics

4:29

of concern to audiences. Hi

4:31

Andy. Hello. So Tomoko let's

4:34

start with you. What kind of things do companies

4:36

need to be worried about when they're thinking about

4:38

their AI or perhaps buying an AI product for

4:40

the business? So I think we

4:42

have to consider the actual software development

4:45

lifecycle. When we actually even start by

4:47

thinking of developing an AI product the

4:49

question is this the first in the

4:51

lifecycle development you have to think about

4:53

how the product is designed, you have

4:56

to look at the data because the

4:58

data in itself could be inherently biased.

5:01

So that's one. The second one is once

5:03

it's being developed we have some

5:05

statistics that say that although a

5:07

lot of companies say that they're

5:09

actually implementing AI ethics within their

5:12

companies, the people who are actually

5:14

developing these AI driven products they

5:16

don't know how to translate those

5:18

principles into practice. So

5:20

when they're actually starting to code some of these

5:23

products, how are these principles

5:25

actually being translated in the day-to-day

5:27

development of this software? And

5:29

then third the question of course is when

5:31

you go into testing of the products, you

5:34

know it's in a rather controlled environment.

5:36

So once then the product gets launched

5:38

out into the world the question then

5:40

is who's monitoring it and how can

5:42

we make sure that what was developed

5:44

within the lab doesn't get the bias

5:46

that they didn't expect it to be

5:48

and also the fact that people

5:50

are going to go to use it in the

5:53

right way as well. So these are some of

5:55

the sort of life cycle approach issues that you

5:57

may have to look at. So we

5:59

need to really be thinking throughout that life cycle

6:01

when we're building, testing and using how does

6:03

the ethical framework fit into every aspect of

6:06

it? That's correct. And Andy, turning to you,

6:08

tell us a bit more about Acumen. What

6:10

was it established to do and what kind

6:12

of services are you providing there? Well

6:15

the founder Paul Howarth about 14 years

6:17

ago realised there was a gap in

6:19

capability to take advantage of

6:21

human feedback in the generation of

6:23

insights. So there's lots of capability

6:25

to understand, score this out of

6:28

one to five, but when someone

6:30

actually gives a dialogue response,

6:32

like a feedback review on Amazon for

6:34

example, then there wasn't the capability to

6:36

do anything with that. So he developed

6:38

a capability with his team. It's used

6:41

to basically turn any time when humans

6:43

are talking to identify to

6:45

a granular level key aspects of the

6:47

meaning in the context such as topics

6:50

and emotions in addition to sentiment, two

6:52

very different things. Sentiment is basically a

6:54

metric of opinion, positive, negative, neutral, whereas

6:56

emotions of which there are many are

6:59

actually one of the key drivers of

7:01

behaviour and it can be used anywhere

7:03

where those conversations are going on. So

7:05

for example in the fast moving consumer

7:07

goods space, lots of

7:10

application in health care, in mental

7:12

health. So yeah it's pretty unlimited

7:14

really. Great, so while this is AI

7:16

that's drawing conclusions about people using that data,

7:18

what you're saying is it's doing it differently

7:20

from the tech that everyone's currently talking about.

7:22

There's a huge amount of hype going on

7:25

in the minutes around generative AI, so chat

7:27

GPT people will know about. The type of

7:29

AI that Acumen has been doing

7:31

for the last 15 years is very different,

7:33

it's rule-based. The simple way of looking at

7:35

it is it's human created and curated and

7:38

rather than being a statistical model

7:40

like large language models are, it's

7:42

completely transparent. So because humans subjectively

7:44

have said well this bit of

7:46

human language means this, for the

7:48

first time the machine, it's immensely

7:50

powerful over numbers, hasn't got a

7:52

clue, bless it, over what humans are

7:54

banging on about normally. So if you can give

7:57

the machine that information it can then use all

7:59

its processing computations. I

10:01

think that the lack of transparency over

10:03

the large language models is absolutely a

10:06

rife area of risk for ethical

10:08

and biased responses. The move towards

10:11

retrieval augmented generative AI when you're

10:13

using those more sort of proofed

10:15

data sets is absolutely the way

10:18

forward and that's fundamentally what a

10:20

rule-based model is. So Tomoko,

10:22

you're focusing on suppliers who put ethics

10:24

front and centre going to become increasingly

10:27

important in this idea of digital responsibility

10:29

for companies. Yes, and I

10:31

think we see it in manufacturers

10:34

first. One company that we

10:36

interviewed was Deutsche Telekom, which has been

10:38

a pioneer and a forefront in the

10:40

AI ethics world. And they

10:42

actually have gone ahead and trained all

10:44

of their company employees, which is a

10:47

rather standard way of making

10:49

sure that AI ethics is distributed

10:51

within the organisation. And

10:54

at the same time though, I think they

10:56

are also going ahead and making sure that

10:58

all of their contractors and their supply chain

11:00

are adhering to their AI ethics policy.

11:03

So it's not only within the boundaries

11:05

of the company itself, but it's going

11:07

beyond. And I think that's what's important

11:09

when it comes to AI ethics, because

11:11

when a software or a product is

11:13

developed, it's not only developed within the

11:15

confines of a company. Absolutely. I think

11:17

you were looking at some study that

11:19

showed that there's many, many charters out

11:22

there. So how do businesses go about navigating

11:24

all the different documents and rules

11:26

that are out there already? That

11:28

is true. I think there's about 200, over 250 principles

11:30

and charters out there for AI

11:34

ethics. And that is now

11:36

the conversation really is getting into

11:38

the operationalisation of AI ethics and

11:41

what can companies do about it.

11:43

I think with any type of

11:45

initiative, it is about execution. I

11:47

think there are these codified types

11:49

of mechanisms that are helpful, but

11:52

at the end of the day, these codified

11:54

mechanisms need to start changing our behaviour. So

11:57

I think the real question is in the long term, how

11:59

can we live these... principles and

12:01

ideals. Many companies

12:03

have trainings at regular basis. Some

12:06

companies have been looking at AI

12:08

ethics advisory boards. Some

12:10

of these companies don't actually have the

12:12

intelligence or the knowledge about how to

12:15

do AI ethics within the company. So

12:17

external experts can help. But I think

12:19

there is a case to be made

12:21

for each individual taking initiative and responsibility

12:23

themselves to be able to sort of

12:25

guide and coach each other. We need

12:27

to have some core principles that I

12:29

think we stick to and then if everybody is

12:32

able to do that maybe as a collective we'll

12:34

be able to have an impact and to make

12:36

sure that we're responsible as a whole organization. I

12:38

think the whole discussion about AI ethics

12:40

is whether you see it as a

12:43

value or whether you see it as

12:45

compliance. And if you take the

12:47

compliance approach of course many things tend

12:49

to be cost driven or risk driven.

12:51

And one of the things

12:53

that we'd like to emphasize is that

12:55

AI ethics could also be of value

12:57

to companies. It could be of competitive

12:59

advantage and those companies who commit and

13:03

demonstrate that they are taking

13:05

privacy seriously, that they're taking

13:07

digital responsibly seriously. A lot

13:10

of the customers like that and so think of

13:12

it as a value generating approach rather than

13:14

a risk or a compliance approach. By

13:17

comparison I use is with the health and safety

13:19

industry or within a company. You have

13:21

the health and safety director or health

13:23

and safety manager, one person that's their

13:25

responsibility. In my experience that's not going

13:27

to make a real change to the

13:29

company unless everyone in the company is

13:31

educated about the importance of health and

13:33

safety but also more importantly they get

13:35

it. They're actively looking, they understand why

13:38

it's important and then as Tamaka was

13:40

saying that it absolutely does drive productivity

13:42

and therefore revenue and your

13:44

bottom line. Absolutely a lot of risk

13:46

and reward for businesses. And Andy I recently

13:48

read a piece and it was about how

13:51

AI automatic translations it was jeopardizing

13:53

people's asylum claims so before they would

13:55

have struggled to get someone to translate

13:57

but because there was misreading and misinformation

13:59

in the... translation, it was really not

14:01

helping their cases. But

14:03

you said that your approach allows for this

14:05

genuine adjustment to biases. Tell us a bit

14:08

more about that. So ChatGPT,

14:10

for example, when being asked how many

14:12

emotions there are in the human experience

14:14

has come back to us with numbers

14:17

as large as 138,000, which may or may not be true, but

14:21

clearly is not very practical when you're trying to

14:23

understand the behaviour drivers

14:25

of humans. So our platform has 19

14:27

to 22 now actually, I think. And

14:29

so that gives you the full

14:32

gambit that's relevant to understanding what

14:34

might be driving human behaviour. And

14:37

this is particularly relevant, for example, within

14:39

the NHS, which is the UK's National

14:42

Health Service, where we deliver through our

14:44

partners, Civico, into lots of trust to

14:46

help them understand the patient experience and

14:48

the practitioner experience. If you can understand

14:50

the topics they're talking about, but also

14:52

the emotions around those topics, and for

14:54

example, frustration or delight, then you can

14:57

very quickly extract very reliable insights, which

14:59

can inform your plan. And at any

15:01

point, if anyone questions things, you can

15:03

open the box and go, well, here

15:05

you go. This is why the systems

15:07

highlighted this. And what do you think

15:09

about that? Do you not like that? Well, then we can

15:11

modify it. But we've been doing it for so long that

15:14

it's actually a very attuned model. But again,

15:16

it's opening that transparent box. So you know

15:18

the workings of how you came to that

15:20

analysis. So Tomiko, how important is

15:22

it to a company's reputation that they

15:24

follow ethical AI practices? I mean, would

15:26

that give them a competitive advantage? We

15:29

don't have specific data to say

15:32

that it gives competitive advantage. But

15:34

there are some surveys, especially related

15:36

to digital trust, the consumers are

15:38

willing to evaluate companies who do

15:40

and commit to these types of

15:43

principles as just being higher. Stakeholders

15:45

or shareholders are willing to invest in more

15:47

of these responsible and ethical companies and

15:50

consumers are willing to buy their products.

15:52

So there is some disparate data that

15:54

shows that there is value to doing

15:57

this. And on the other side is the risk, as

15:59

we talked about, the risk of the pandemic. about greenwashing when we

16:01

talk about environmental. Are there aspects

16:03

of this that could spill over

16:05

into tech and AI? Of course

16:07

there's already a term that's called

16:09

machine washing. So it's already here.

16:11

I think what companies really need

16:13

to be diligent about is

16:16

that when they make these public

16:18

commitments that they go ahead and

16:20

execute and operationalize. Organizational change, embedding

16:22

new practices is a very, very

16:25

difficult endeavor. So one of the

16:27

things I would just recommend is

16:29

that if you have a

16:31

grand goal in terms

16:33

of committing to AI ethics or

16:36

digital responsibility, divide it into sub-goals

16:38

that are tangible and it can

16:40

easily be executed by the teams or

16:42

by the functions or by divisions. And

16:45

then the accumulation of these sub-goals will

16:47

turn into results and results that people

16:50

can see. And I think that's the

16:52

most important things that organizations should be

16:54

doing right now. And Andy,

16:56

you supply these different kinds of AI services

16:58

to clients. Are you seeing this sort of

17:01

consideration around safety and governance in the clients

17:03

you work with? I think certainly,

17:05

but specifically our experience is to health

17:07

care and mental health, where and hallucination

17:09

as errors in large language models are

17:12

jocularly referred to, that's very critical. That's

17:14

life-threatening. So you can't have those. So

17:16

at any time when you're dealing with

17:18

anyone's PII in any space, it should

17:21

obviously be sacrosanct and hugely respected, but

17:23

no more so than in health care.

17:25

It's nothing new to us, but I

17:27

think as people start leaning more on

17:30

the generative AI front, then not being

17:32

100% confident what the response might

17:36

be in a given situation is

17:38

leaving the door open for

17:40

accidents and challenges and

17:43

ultimately a drop in share price.

17:45

And do you think the management boards

17:47

and the C-switch should appoint someone or

17:49

is it about educating everyone about how

17:51

to use these ethical frameworks? It's

17:53

a really good question. I think that's the the knob

17:56

of it all. I think all of the above, as

17:58

you mentioned, Susie, because let's not forget AI. is

18:01

not just at work, it's flopped over

18:03

into our entire existence. So it's almost

18:05

you should make sure that you know

18:07

what's going on as an individual and then everyone's going

18:09

to care to a certain extent within

18:11

the corporate structure but it should

18:13

be a multi-tier approach and understanding.

18:15

A whole tier from the individual to

18:17

the business. Absolutely. And Tomoko, you're part of the

18:19

IMD Business School and it's turning out those top

18:22

executives of the future they're going to be in

18:24

the C suite in years to come. So what

18:26

are you telling the people who pass through the

18:28

school about these types of issues and what's the

18:30

next things that they should be thinking about in

18:32

their careers? We are trying

18:35

to tell people that everybody has

18:37

a responsibility. They have responsibility to

18:39

these issues that go beyond the

18:41

company themselves and that

18:43

this type of responsibility is something that one

18:45

needs to be aware of but also the

18:48

fact that they have a responsibility to be

18:50

able to coach and make the

18:52

other people within their team to be aware of

18:55

it as well. So I think that's the first

18:57

thing. The second thing is

18:59

what type of organizations do we want to

19:01

work with and do these executives want to

19:03

build? Is it an organization that

19:05

can handle these multiple goals that we have

19:07

to handle which is not only profit

19:10

but also some of these social,

19:12

environmental and ethical issues? And

19:14

I think the third one very much is as we

19:16

try and coach people to be

19:18

able to handle multiple goals within

19:21

their workplace. How can you do

19:23

that in a way that is effective for the individual

19:25

and that adheres to their value and at

19:27

the same time is also good for the planet?

19:29

So I think those are the three areas

19:32

that we'd like executives to walk away with.

19:34

Thinking about the future also for you Andy,

19:36

what do you think is coming down the

19:38

line to suppliers like yourself and what kind

19:40

of legislation or requirements are going to be

19:42

placed on you? I think it's fair to say

19:44

to a certain extent in the world of AI as

19:47

of today if anyone tells you that they know what's

19:49

going to happen in the next six months or beyond

19:51

I'd highly question that. I think there's a lot of

19:53

fear, there's a lot of lobbying going on in different

19:55

areas and so I'd go

19:57

slow because if your AI driven capabilities

20:00

falls down for a failure in delivering

20:02

capability but also being roused for being

20:04

non-compliant ethically or otherwise it's going to

20:06

be very damaging to you. There is

20:08

absolutely a need for regulation I find

20:10

it personally quite interesting to see

20:12

how you regulate something that is not particularly

20:14

definable and morphs as Tomoko has said very

20:17

quickly. We have to try and do our

20:19

best to protect those who need protecting so

20:21

I think you're going to have to be

20:23

very agile and you're going to have to

20:25

understand the risks almost as they

20:27

occur. Absolutely and we like to

20:29

end our interviews by offering a final

20:31

nugget so that business executive audience who

20:34

perhaps haven't done anything yet around AI

20:36

ethics they've maybe messed around with generative

20:38

AI or they bought some tools but

20:40

they don't have that board of ethics

20:42

or that charter yet. What

20:44

would you say is the first thing that

20:46

they should be thinking about doing today? Let's

20:48

turn to you first Tomoko. Well

20:50

as a mindset please keep

20:53

in mind that the analog world and the digital

20:55

world is the same so whatever

20:57

values you have in the analog world

20:59

should be extended to the digital world

21:01

it's not separate. Absolutely and

21:04

Andy what would you say is your

21:06

golden nugget of insight? Well I'm

21:08

going to contradict myself as always having

21:10

just said that you should get your

21:12

entire team conversant with AI definitely do

21:14

that but also have a well-informed friend

21:16

to call someone who lives and breathes

21:18

this stuff so when there are changes

21:20

they understand it first and they understand

21:22

its implications and have them on speed

21:24

dial. A

21:30

big thank you to Tomoko, Yokoi and

21:32

Andy Crouch for sharing their insights on

21:34

AI and its ethical challenges. If

21:38

you're enjoying these kinds of insights

21:40

you really need to be reading

21:42

secure futures. Kaspersky's digital magazine about

21:44

all kinds of tech trends for the

21:46

discerning business executives. We've

21:48

got an opinion piece about regulating

21:50

chat GPT, one about emotionally intelligent

21:53

AI for recruitment and articles from

21:55

this season's insight story including making

21:57

the most from generative AI. and

22:00

an intro to digital sovereignty. You

22:02

can find the link to Secure Futures in the

22:05

Insight Stories show notes. Everyone

22:10

agrees that AI and machine learning

22:12

offer incredibly exciting opportunities for businesses.

22:15

But as we've been saying, they have to

22:17

be used ethically and safely. To

22:19

give us some insight into how to make the

22:21

most of this tech while remaining secure, I'm

22:24

joined by Dr. Armin Hazbini, Head

22:26

of Research Centre, Middle East, Turkey

22:29

and Africa, to Kaspersky's global research

22:31

and analysis team known as Great.

22:34

So, Armin, what are the key things that businesses should

22:36

be thinking about if they're going into this area? When

22:38

we talk about AI, a lot of us

22:40

do not realise, I think, that AI

22:43

is not going to be ethical by

22:45

itself. It's not

22:47

going to self-define its ethical

22:49

standards. We also need

22:51

to agree that when AIs that

22:54

are being used everywhere around

22:56

the world today, they are

22:58

programmed with ethical standards. And these

23:00

ethical standards, they

23:02

do not represent everyone. They usually

23:04

represent a biased definition of

23:07

ethics that is

23:09

specific to a number of people,

23:11

maybe programmers, developers, maybe an organisation,

23:13

maybe the country within which

23:16

the organisation operates. I

23:18

think what is also important to say

23:20

here is that there is

23:22

almost no way the public

23:25

is able to check, evaluate,

23:28

critique or enhance

23:31

these ethical capabilities within

23:33

this AI engine, within the

23:36

AI algorithms that

23:38

we are all using nowadays, which

23:40

is a major challenge, of course. So,

23:43

do you think there's a role there for

23:45

regulation of these

23:47

technologies, because the business itself can't

23:50

really fully evaluate the language models

23:52

and the inputs? Yes, indeed.

23:54

We need security and safety by design

23:57

and then continuous verification of

23:59

this. security and safety by design.

24:02

Such would require transparency measures,

24:05

especially by big tech vendors

24:07

and allowing the public to

24:10

influence such technologies and their

24:12

development. In reality, we're

24:14

asking the public to adopt technologies that can do

24:16

a lot of damage without

24:18

giving that same public the

24:21

capabilities to make sure that

24:23

damage will not happen. We

24:26

did the same before with social

24:28

media and social media technologies and

24:30

we've seen, sadly,

24:32

a lot of incidents such as

24:34

data leaks, abuse of data, scandals

24:37

of fake news, etc. Do

24:39

you think we're in danger of

24:41

repeating those same patterns unless regulation

24:43

happens, maybe more of a regional

24:46

level, for example, in the European

24:48

Union? On the level of the European Union, I

24:50

think regulations are moving

24:52

very fast into the areas

24:55

of artificial intelligence. Still, I

24:58

think AI could

25:00

cause a lot of damage. It's

25:02

much more dangerous than social media

25:05

and we definitely need these rules and

25:07

laws in place as quickly as possible.

25:10

What about the data sets themselves? If

25:12

a company is buying a generative AI

25:14

tool, what sort of questions should they

25:17

be asking to make sure that the

25:19

data they put in and also the

25:21

training data that the model is based

25:23

on is actually secure? I

25:25

think an organization will figure out

25:27

by itself what kind of data

25:29

it is allowed

25:31

to put on AI if

25:33

it has the right asset

25:36

management controls in place. Because

25:38

the asset management controls, if they are

25:40

well deployed, they would allow to classify

25:43

the data and they

25:45

would allow to identify which data can

25:47

be available to AI, which data can

25:49

be available to the public, which data

25:51

needs to stay inside

25:54

the organization, confidential, private

25:57

or secret. Thank

26:00

you very much to Amin. Now

26:04

one person who has a lot of thoughts

26:06

about how AI can be better regulated is

26:08

Eugene Kaspersky, a world leading cyber

26:10

security expert who founded Kaspersky in 1997.

26:14

Today he's the CEO at the helm, protecting

26:16

240,000 businesses and 400 million people from cyber

26:18

threats. He's

26:22

a busy man. He recently

26:24

wrote for Kaspersky Daily magazine about how

26:27

he thinks regulators in the industry should

26:29

be thinking about AI to improve trust. Do

26:32

you think he's in favour of heavy regulations,

26:34

night to touch or no regulation? Only

26:36

one way to find out. Check the link in the

26:38

show notes to read it. That's

26:44

it for this edition and in fact

26:46

this series of Insight Story Tech Trends

26:48

Unpacked brought to you by Kaspersky. If

26:51

you've missed any episodes, search for us wherever you

26:54

get your podcasts and you'll find all our other

26:56

shows, jam-packed with all kinds

26:58

of insights for business leaders. We've

27:00

found global experts on topics from

27:02

blockchain to quantum computing and smart

27:04

energy to industrial IoT. Go

27:07

and check it out and if you like it, please

27:09

leave us a rating and give us an excellent

27:11

review and tell us what technologies

27:13

you'd like to hear about in the next season.

27:16

If you want to get ahead, you really can't afford

27:18

to miss it. Till the next

27:20

time, goodbye. Goodbye. Ah, Sir Donley, thank

27:22

you for all the info you shared with us

27:24

this season. It was a pleasure. I'm sure I don't

27:27

even have to ask you this. It was

27:29

absolutely source-wide. Do you know where it came from?

27:32

Have you ever been sick of speaking? Just

27:42

before I go, I wanted to tell you

27:44

about two other great series from Kaspersky that you

27:46

might like. Fast Forward

27:48

by Tomorrow Unlocked explores the past, present

27:50

and future of the technologies around us.

27:53

Season 2 features six fresh new

27:55

episodes including augmented body technology, how

27:58

tech is changing family life. and

28:00

women in gaming. Plus, if you want

28:02

to hear the latest news and views

28:04

from the world of cybersecurity, join Jeff

28:06

Esposito in the US and David Buxton

28:09

in the UK for Kaspersky Transatlantic. They

28:11

chat security around current tech news and have

28:13

the lowdown on the latest data breaches. You'll

28:16

find links to both these series in our show notes,

28:18

but you'll also find them wherever you get your podcasts.

28:21

So track them down and click follow so you don't

28:23

miss an episode.