July 13, 2026

Cyber Security - In the Age of AI

Cyber Security - In the Age of AI
Apple Podcasts podcast player icon Spotify podcast player icon Spreaker podcast player icon iHeartRadio podcast player icon Amazon Music podcast player icon Castbox podcast player icon Pandora podcast player icon Deezer podcast player icon Podcast Addict podcast player icon Podchaser podcast player icon Youtube Music podcast player icon Castamatic podcast player icon Castro podcast player icon Fountain podcast player icon RSS Feed podcast player icon Goodpods podcast player icon Overcast podcast player icon PocketCasts podcast player icon Podurama podcast player icon Podverse podcast player icon Podyssey podcast player icon
Apple Podcasts podcast player icon Spotify podcast player icon Spreaker podcast player icon iHeartRadio podcast player icon Amazon Music podcast player icon Castbox podcast player icon Pandora podcast player icon Deezer podcast player icon Podcast Addict podcast player icon Podchaser podcast player icon Youtube Music podcast player icon Castamatic podcast player icon Castro podcast player icon Fountain podcast player icon RSS Feed podcast player icon Goodpods podcast player icon Overcast podcast player icon PocketCasts podcast player icon Podurama podcast player icon Podverse podcast player icon Podyssey podcast player icon
Season 2

Cyber threats are becoming more sophisticated, frequent, and personal. AI is creating both powerful defenses and powerful new risks. Welcome to the podcast. Today we’re diving into one of the most urgent issues in business and society right now, Cybersecurity. Now artificial intelligence is raising the stakes even higher. AI is helping defenders detect threats faster, but it’s also helping attackers create more convincing scams, automate attacks, and exploit human trust in new processing ways.

ASharpe Outlook is broadcast live Mondays at 8AM PT on K4HD Radio - Hollywood Talk Radio (www.k4hd.com) part of Talk 4 Radio (www.talk4radio.com) on the Talk 4 Media Network (www.talk4media.com). ASharpe Outlook TV Show is viewed on Talk 4 TV (www.talk4tv.com).

ASharpe Outlook Podcast is also available on Talk 4 Media (www.talk4media.com), Talk 4 Podcasting (www.talk4podcasting.com), iHeartRadio, Amazon Music, Pandora, Spotify, Audible, and over 100 other podcast outlets.

Become a supporter of this podcast: https://www.spreaker.com/podcast/asharpe-outlook--6643730/support.

WEBVTT

1
00:00:01.000 --> 00:00:04.559
This program is designed to provide general information with regards

2
00:00:04.599 --> 00:00:07.559
to the subject matters covered. This information is given with

3
00:00:07.639 --> 00:00:12.000
the understanding that neither the hosts, guests, sponsors, or station

4
00:00:12.199 --> 00:00:19.039
are engaged in rendering any specific and personal medical, financial, legal, counseling,

5
00:00:19.280 --> 00:00:23.440
professional service, or any advice. You should seek the services

6
00:00:23.480 --> 00:00:29.559
of competent professionals before applying or trying any suggested ideas.

7
00:00:31.760 --> 00:00:34.560
Hello and thank you for tuning in to a Shark

8
00:00:34.640 --> 00:00:38.880
Outlook on pay for HD radio and talk or TV.

9
00:00:39.640 --> 00:00:43.079
Hi am Angela Sharp your host. I'm on to your

10
00:00:43.119 --> 00:00:47.240
discussions with industry experts will give you the steps, tools

11
00:00:47.280 --> 00:00:51.359
and information to be successful in business and to prepare

12
00:00:51.399 --> 00:00:52.000
you to be.

13
00:00:52.159 --> 00:00:56.799
Your best self. Hello and thank you for joining us

14
00:00:56.799 --> 00:01:00.600
at a Sharp Outlook. I'm really excited about this topic

15
00:01:00.719 --> 00:01:04.719
because I experience a lot of questions while I'm doing

16
00:01:04.959 --> 00:01:09.640
my work in finance, and so we're going to get

17
00:01:09.680 --> 00:01:14.640
into it. Cyber threats are becoming more sophisticated, frequent, and

18
00:01:14.760 --> 00:01:19.640
more personal. At the same time, AI is creating both

19
00:01:19.959 --> 00:01:27.120
powerful defenses and powerful new risk Welcome to the podcast. Today,

20
00:01:27.159 --> 00:01:29.920
We're going to dive into one of the most urgent

21
00:01:30.079 --> 00:01:35.560
issues in business and society, right now cybersecurity in the

22
00:01:35.680 --> 00:01:39.840
age of AI. We live in a world where almost

23
00:01:39.879 --> 00:01:49.040
everything runs on connected systems, our finances, healthcare, communications, businesses, infrastructure.

24
00:01:49.719 --> 00:01:54.120
That means cybersecurity is no longer just an it concern.

25
00:01:55.920 --> 00:02:00.879
It's a how should I say it? Trust issue, a

26
00:02:00.920 --> 00:02:05.959
business issue, and in many cases, a public safety issue.

27
00:02:07.000 --> 00:02:11.599
And now artificial intelligence is raising the stakes even higher.

28
00:02:12.479 --> 00:02:18.439
AI is helping defenders to detect threats faster, and that's wonderful,

29
00:02:18.639 --> 00:02:26.439
but it's also helping attackers create more convincing scams, automate attacks,

30
00:02:26.960 --> 00:02:32.159
and exploit human trust in brand new ways. So Detay's

31
00:02:32.240 --> 00:02:37.400
conversation is about what's really happening in cybersecurity right now,

32
00:02:37.719 --> 00:02:42.919
when businesses are still getting wrong, and how AI is

33
00:02:43.039 --> 00:02:47.719
changing both the threat and the response. I'm joined by

34
00:02:48.960 --> 00:02:53.000
a leader in the cybersecurity space who works directly with

35
00:02:53.240 --> 00:02:59.000
organizations trying to protect themselves in rapidly changing environments. We're

36
00:02:59.000 --> 00:03:02.319
going to go beyond the basics today. We'll talk about

37
00:03:02.439 --> 00:03:07.560
real risk, hard truths, leadership failures, and what companies need

38
00:03:07.639 --> 00:03:12.159
to understand before they become the next headline. We'll talk

39
00:03:12.159 --> 00:03:15.919
about today's biggest threats, how companies should think about cyber

40
00:03:16.000 --> 00:03:19.840
risk and what the rise of AI means for the

41
00:03:19.879 --> 00:03:25.599
future of security. My guest today is Doug Simmons and

42
00:03:25.759 --> 00:03:31.639
Genevieve and we are there from Intelligent Forensics x Inc.

43
00:03:32.560 --> 00:03:35.400
And Genevieve and Doug have been guested on the show

44
00:03:35.439 --> 00:03:39.759
before and you can find their previous podcast on a

45
00:03:39.879 --> 00:03:43.719
Sharp Outlook. You can search on in your podcast apps

46
00:03:44.120 --> 00:03:50.280
wherever you are and search for a Sharp Outlook. I'd

47
00:03:50.360 --> 00:03:53.479
like to invite Doug and Genevieve to the show so

48
00:03:53.520 --> 00:03:56.800
that we can get into this discussion because I tell

49
00:03:56.879 --> 00:04:01.639
you this can be a areas issue and we need

50
00:04:01.680 --> 00:04:07.639
experts like them in their helping industries and companies be

51
00:04:07.759 --> 00:04:11.759
aware of what they're facing, because I don't think they're

52
00:04:11.919 --> 00:04:14.960
they're aware. They're just all tickled about the AI and

53
00:04:15.159 --> 00:04:17.120
getting to see the little box and see the little

54
00:04:17.120 --> 00:04:19.800
pictures and all the little things. But there's a lot

55
00:04:19.879 --> 00:04:24.959
more too what's going on. So good morning you guys,

56
00:04:26.199 --> 00:04:30.920
Good morning, good morning. So before we start, if you can't,

57
00:04:31.000 --> 00:04:34.600
just give a brief description about yourselves and then we're

58
00:04:34.600 --> 00:04:36.120
going to get into this topic.

59
00:04:37.879 --> 00:04:38.199
To the.

60
00:04:40.240 --> 00:04:40.399
HI.

61
00:04:40.519 --> 00:04:45.399
So my name is is Genevieve? Am I coming through? Okay? Yes, okay, great.

62
00:04:46.480 --> 00:04:50.800
So I started intelligent computer forensics way back.

63
00:04:50.639 --> 00:04:54.519
When and good morning, good morning, and then.

64
00:04:54.399 --> 00:05:00.600
Just recently i've I've included AI of course, and we

65
00:05:00.720 --> 00:05:04.680
actually are in the process of developing agentic AI with forensics.

66
00:05:05.000 --> 00:05:08.839
So it's a willing, fascinating, interesting world that we live

67
00:05:08.879 --> 00:05:11.439
in these days, and we're really excited to talk about

68
00:05:11.480 --> 00:05:15.160
this topic, this very important topic. And Doug and I

69
00:05:15.199 --> 00:05:18.720
have been working together on this as well, and he's

70
00:05:18.759 --> 00:05:22.439
got some good insights as well. So yeah, I've been

71
00:05:22.480 --> 00:05:25.120
in the industry for a really long time and I'm

72
00:05:25.160 --> 00:05:26.120
just happy to be a board.

73
00:05:26.279 --> 00:05:31.279
Thank you, thank you. Good Yeah.

74
00:05:31.360 --> 00:05:34.120
Yes, my name is security with an IT for a

75
00:05:34.160 --> 00:05:38.319
long time, but a theory stone. Consider me a dinosaur

76
00:05:38.399 --> 00:05:40.040
because I am.

77
00:05:40.720 --> 00:05:42.720
I have a lot involved in.

78
00:05:44.560 --> 00:05:47.720
The deployment to roll out and the word I'm going

79
00:05:47.759 --> 00:05:52.199
to use many times in today's discussion governance.

80
00:05:53.680 --> 00:05:54.160
Of AI.

81
00:05:55.240 --> 00:05:57.800
And I think that when we get into this, we'll

82
00:05:57.839 --> 00:06:01.560
see just how important governance is now in an IT environment,

83
00:06:01.600 --> 00:06:07.920
specifically cybersecurity, protecting information assets, protecting the people of the organization,

84
00:06:08.040 --> 00:06:11.360
the customers. We knew it was coming for decades. It's

85
00:06:11.480 --> 00:06:14.360
that we need to discuss today.

86
00:06:15.000 --> 00:06:18.560
Yeah, that is that is really important. You know, with

87
00:06:18.639 --> 00:06:22.279
my finance background and tech background, I do believe in

88
00:06:22.439 --> 00:06:30.000
governance and regulations and processes and policies and oh not

89
00:06:30.160 --> 00:06:33.439
just having them in a book, but actually following them

90
00:06:33.560 --> 00:06:37.800
and implementing them. So that's a really important you know,

91
00:06:38.160 --> 00:06:41.040
and so I'm excited to get into this. So how

92
00:06:41.199 --> 00:06:46.040
is AI changing cybersecurity from your perspectives?

93
00:06:47.279 --> 00:06:51.800
You want to go ahead, Doug, Let's go with Jenevie first,

94
00:06:51.920 --> 00:06:55.720
and okay, if you want to do it that way.

95
00:06:56.279 --> 00:06:56.920
Okay, sure.

96
00:06:57.800 --> 00:07:02.480
So the landscape of AI is changing cybersecurity in many

97
00:07:02.480 --> 00:07:06.680
different ways. People are using it, they're implementing it. However,

98
00:07:07.160 --> 00:07:09.360
I think what's really we really need to stress here

99
00:07:09.480 --> 00:07:12.959
is that they are not put in the proper safeguards

100
00:07:13.360 --> 00:07:15.639
and the guardrails that need to be put in, and

101
00:07:15.680 --> 00:07:19.240
as the results, there's maliciousness.

102
00:07:18.519 --> 00:07:19.439
That's coming through.

103
00:07:20.079 --> 00:07:23.360
And a lot of oftentimes the maliciousness happens through the

104
00:07:23.439 --> 00:07:28.800
API connections or the MCP connections that needs a gentic

105
00:07:28.839 --> 00:07:34.639
AI agents are connected to. So we still, even though

106
00:07:34.680 --> 00:07:37.800
we're using all these different tools, we always have to

107
00:07:37.879 --> 00:07:42.040
keep in mind security top of mind. Otherwise we're opening

108
00:07:42.120 --> 00:07:49.319
ourselves to the adversary potentially being able to either sever

109
00:07:49.439 --> 00:07:52.399
our connection or come in through our connections somewhere or another.

110
00:07:53.160 --> 00:07:56.519
All the same hacks that are out there that have

111
00:07:56.639 --> 00:07:58.680
been out there are still apply to AI.

112
00:07:59.279 --> 00:08:01.000
So it's great to use it, but we teed to

113
00:08:01.079 --> 00:08:01.759
use it safely.

114
00:08:02.759 --> 00:08:08.439
Yeah, Yeah, definitely, what do you say done?

115
00:08:09.120 --> 00:08:14.399
Yeah, the cybersecurity dramatically and for the right reasons, I

116
00:08:14.399 --> 00:08:17.720
would say, over the past three years. First and foremost,

117
00:08:18.040 --> 00:08:27.360
like ourselves required to monitor and determine and reacting, detecting,

118
00:08:28.279 --> 00:08:34.639
stopping behavior from occurring that would lead to the aggressive information,

119
00:08:35.000 --> 00:08:39.039
the automated manufacturing system, et cetera. There's a lot of

120
00:08:39.080 --> 00:08:44.159
things that cybersecurity does. Using AI within cybersecurity is a

121
00:08:44.200 --> 00:08:48.559
wonderful thing because theoretically, and I'll always put it, theoretically,

122
00:08:49.480 --> 00:08:52.840
you're allowed to do many more things far faster when

123
00:08:52.879 --> 00:08:56.039
you have AI inside your tool helping you, right, and

124
00:08:56.039 --> 00:08:59.200
the keyword is helping you because think of it as

125
00:08:59.279 --> 00:09:02.600
AI as a million little people now in your tool

126
00:09:02.840 --> 00:09:05.639
that are well trained and are able to scrub and

127
00:09:05.759 --> 00:09:08.120
look at all of the things, detect you know, what

128
00:09:08.279 --> 00:09:12.759
is happening, who you know, what systems are talking to,

129
00:09:12.799 --> 00:09:15.399
what data is being transferred. Who's trying to come in

130
00:09:16.200 --> 00:09:19.200
from the outside, what's trying to go out from the inside.

131
00:09:19.240 --> 00:09:23.440
All these things our cybersecurity tools traditionally look at and

132
00:09:23.519 --> 00:09:27.000
AI AI is now helping these tools look at these

133
00:09:27.080 --> 00:09:31.080
things and detect far more rapidly and far more pervasively.

134
00:09:31.360 --> 00:09:34.159
That's a wonderful thing. However, at the end of the day,

135
00:09:36.720 --> 00:09:39.840
we still need human beings to be able to determine

136
00:09:40.279 --> 00:09:43.399
what should be done based on the findings by the tooling.

137
00:09:43.720 --> 00:09:46.480
It will always be that way. And so I'll stop

138
00:09:49.919 --> 00:09:52.480
this particular proportion of my discussion saying that it is

139
00:09:52.559 --> 00:09:54.039
incumbent upon panacea.

140
00:09:54.279 --> 00:09:57.519
It doesn't do it for you.

141
00:09:57.840 --> 00:10:02.919
It helps you, but you human being, the organization and

142
00:10:03.000 --> 00:10:08.000
comprised of human beings, are responsible for analyzing the results. Today,

143
00:10:08.039 --> 00:10:13.759
I find it's you analyzing the recommendations that AI make

144
00:10:13.879 --> 00:10:17.000
based on what it finds, etc. So, to go back

145
00:10:17.039 --> 00:10:20.639
to what Genevieve was saying, to be the be all

146
00:10:20.759 --> 00:10:23.799
end all to do it all, so we can just

147
00:10:23.840 --> 00:10:25.240
sit up, put our feet up on the desk and

148
00:10:25.279 --> 00:10:27.679
say things are running smoothly. That is a very very

149
00:10:27.759 --> 00:10:34.279
dangerous approach, and it's coming to surface now. And as

150
00:10:34.320 --> 00:10:38.759
we get into this discussion further, I'll have some i'll

151
00:10:38.759 --> 00:10:42.000
call case studies or examples where companies found out the

152
00:10:42.039 --> 00:10:44.559
hard way that you cannot just rely on AI to.

153
00:10:44.559 --> 00:10:45.600
Do it for you.

154
00:10:45.159 --> 00:10:47.720
You have to be the person, the man in the

155
00:10:47.720 --> 00:10:49.600
middle of the person in the middle who's looking at

156
00:10:49.600 --> 00:10:53.320
what's going on and allowing it to continue or going

157
00:10:53.320 --> 00:10:54.200
down a different path.

158
00:10:54.320 --> 00:10:55.480
So I'll stop there, thank you.

159
00:10:55.639 --> 00:10:55.759
Yes.

160
00:10:56.399 --> 00:10:56.639
Yeah.

161
00:10:56.679 --> 00:10:58.480
In fact, that's one of the things I've been finding

162
00:10:58.480 --> 00:11:03.919
out when I'm talking at a company. They must have

163
00:11:04.000 --> 00:11:06.960
expected it to be able to just you know, place

164
00:11:07.519 --> 00:11:10.840
human beings and just come up with, you know, solutions.

165
00:11:11.399 --> 00:11:16.480
But as I was mentioning, if you have bad processes

166
00:11:16.519 --> 00:11:18.679
and you don't have good policies, and you don't have

167
00:11:18.960 --> 00:11:23.879
rules and regulations built in that you're following, or you

168
00:11:23.879 --> 00:11:28.919
don't even know what the solution you're expecting is supposed

169
00:11:28.919 --> 00:11:31.799
to be, then you're going to run into some real problems.

170
00:11:31.799 --> 00:11:34.879
And that's what I'm seeing that they're just thinking they

171
00:11:34.879 --> 00:11:39.960
can just just turn on AI and just no human intervention,

172
00:11:40.679 --> 00:11:47.159
no people sitting there looking at it and actually analyzing

173
00:11:47.200 --> 00:11:50.399
what the output is. They're just taking it for granted

174
00:11:50.679 --> 00:11:53.759
that it's accurate and boy, it is blowing up some

175
00:11:53.840 --> 00:11:58.159
real egos right now. There are companies having to you know, say,

176
00:11:58.159 --> 00:12:00.279
you know what, all the stuff I sent you just

177
00:12:00.960 --> 00:12:04.679
it's all wrong. So now you've lost your reputation and

178
00:12:04.759 --> 00:12:09.480
people's trust and and you know, it's unfortunate that they

179
00:12:09.480 --> 00:12:13.039
didn't think that through before they even did this. But

180
00:12:13.480 --> 00:12:15.799
that's why we're here to try to give them some

181
00:12:15.799 --> 00:12:19.759
some guidance and some help and give them some resources

182
00:12:19.840 --> 00:12:24.039
like intelligent forensics SECS to go out and help them.

183
00:12:24.080 --> 00:12:30.519
Because I'm there having to deal with the cleanup. When

184
00:12:30.559 --> 00:12:32.679
I go in there, you know, we have a disaster

185
00:12:32.919 --> 00:12:36.240
that has occurred because they just thought he I was

186
00:12:36.279 --> 00:12:39.440
just gonna, you know, build their profit line up by

187
00:12:39.480 --> 00:12:42.519
getting rid of people, and that's not it. So you know,

188
00:12:44.639 --> 00:12:46.960
what are some of the biggest ways AI is helping

189
00:12:47.279 --> 00:12:48.559
defenders right now?

190
00:12:48.720 --> 00:12:49.639
It's not that.

191
00:12:51.320 --> 00:12:51.679
M hm.

192
00:12:52.600 --> 00:12:54.440
So I'm gonna I'm gonna go ahead and take over

193
00:12:54.480 --> 00:12:58.399
here for Defenders's you.

194
00:12:58.480 --> 00:12:59.440
Hit the nail on the head.

195
00:13:00.840 --> 00:13:06.080
Okay, okay, so what what the way they use it

196
00:13:06.120 --> 00:13:09.919
AI now and the defense If you are part of

197
00:13:10.480 --> 00:13:11.639
the corporation.

198
00:13:11.320 --> 00:13:13.039
Let me just say one thing.

199
00:13:14.039 --> 00:13:17.519
I think you hit the nail on the head that

200
00:13:17.840 --> 00:13:23.519
people are losing their jobs to A. This is not

201
00:13:23.879 --> 00:13:29.200
a sky is full oh calling you know. I'm telling

202
00:13:29.240 --> 00:13:33.759
you I have people I know. I have neighbors who

203
00:13:33.799 --> 00:13:37.559
are losing him lost their jobs.

204
00:13:38.879 --> 00:13:39.600
And their.

205
00:13:40.960 --> 00:13:45.360
Employers told them, because we're placing you.

206
00:13:45.279 --> 00:13:47.279
A, I can do it.

207
00:13:49.879 --> 00:13:53.159
And as you said, Angela, we're starting to find out.

208
00:13:54.919 --> 00:13:56.639
Okay, go ahead.

209
00:13:57.879 --> 00:14:02.679
So to add to that, yes, a lot of corporations

210
00:14:02.799 --> 00:14:05.720
are letting go of their people because they are under

211
00:14:05.759 --> 00:14:09.600
the substances that AI is going to take over their jobs.

212
00:14:09.639 --> 00:14:11.360
It's going to be able to do their job, if

213
00:14:11.399 --> 00:14:13.879
you will, and they're finding.

214
00:14:13.519 --> 00:14:17.840
Out that, yes, there is that to some extent.

215
00:14:18.039 --> 00:14:23.360
However, we still need a person behind the keyboard that's

216
00:14:23.399 --> 00:14:26.639
really understanding what the AI is doing and kind of

217
00:14:26.720 --> 00:14:29.960
keeping track of it, if you will, right, because otherwise

218
00:14:29.960 --> 00:14:32.080
it just kind of it's a wild while West out

219
00:14:32.080 --> 00:14:34.120
there if we just use these AIS and we don't

220
00:14:34.159 --> 00:14:37.080
know what we're doing. What companies are doing is they're

221
00:14:37.080 --> 00:14:40.159
creating AI agents to do their jobs. So they're created

222
00:14:40.200 --> 00:14:43.720
AI AGENTSY for an assistant, and that's how they're using them.

223
00:14:43.720 --> 00:14:47.559
In the defense industry, they're creating an AI assistant that

224
00:14:47.639 --> 00:14:50.840
will actually go out there and be able to gather

225
00:14:51.159 --> 00:14:55.519
the necessary information at a much faster rate than the sims.

226
00:14:56.200 --> 00:15:00.360
You know, the actual security information management systems that are

227
00:15:00.440 --> 00:15:04.039
in place to be able to analyze the traffic a

228
00:15:04.080 --> 00:15:06.720
lot quicker and then have a return back to the

229
00:15:06.759 --> 00:15:09.039
analysts and say, Okay, this is what we have. But

230
00:15:09.080 --> 00:15:11.480
the analyst still has behind the keyboard, it has to

231
00:15:11.559 --> 00:15:15.799
still determined whether or not that's legitimate. Right, So, because

232
00:15:15.799 --> 00:15:18.519
of the hallucinations and the bias and the drifting that

233
00:15:18.559 --> 00:15:21.840
goes on with these agents, we have to ensure that

234
00:15:21.840 --> 00:15:24.720
that's not happening. Now, you can eliminate all of that

235
00:15:24.799 --> 00:15:27.919
when I just spoke on spoke on through the AI

236
00:15:28.000 --> 00:15:29.159
agent if you set it up.

237
00:15:29.039 --> 00:15:29.840
The proper way.

238
00:15:30.080 --> 00:15:32.399
So there is a proper way to do that, right,

239
00:15:32.480 --> 00:15:34.679
So it's going to it actually can go back. It

240
00:15:34.679 --> 00:15:38.639
can go to vector databases, It can do algorithms that

241
00:15:38.720 --> 00:15:42.960
are way beyond what a human can do in a

242
00:15:43.000 --> 00:15:46.120
million seconds of time, and then be able to return

243
00:15:46.320 --> 00:15:49.759
the information that it's needed to the analysts to analyze. Again,

244
00:15:50.279 --> 00:15:53.600
there has to be some a person behind the keyboard

245
00:15:53.679 --> 00:15:59.039
that's actually determining the information is true or not ord

246
00:15:59.279 --> 00:16:02.120
or is you know what it says it is.

247
00:16:02.840 --> 00:16:05.799
Yeah, And that's the key point. They still need that

248
00:16:06.000 --> 00:16:10.279
human intervention with someone who has the knowledge of how

249
00:16:10.320 --> 00:16:13.679
this is supposed to work. Maybe can't do it in milliseconds,

250
00:16:13.919 --> 00:16:16.960
but they able to do it if they were just

251
00:16:17.279 --> 00:16:20.200
handling their job and they know the steps that are

252
00:16:20.240 --> 00:16:23.000
supposed to occur. But they're just throwing these things in

253
00:16:23.039 --> 00:16:26.679
their willy nilly and expect them to be able to

254
00:16:26.759 --> 00:16:31.879
perform miracles. And as I mentioned before, most a lot

255
00:16:31.919 --> 00:16:34.320
of companies, and I know this because they've been working

256
00:16:34.360 --> 00:16:39.279
a long time, they don't even have processes in place

257
00:16:39.320 --> 00:16:43.159
that they follow on a regular basis. So how they

258
00:16:43.279 --> 00:16:48.399
expect for a system to come in and actually get

259
00:16:48.440 --> 00:16:51.320
this all done when you never set up even true

260
00:16:51.399 --> 00:16:54.799
processes of how the information is supposed to flow in

261
00:16:54.840 --> 00:16:58.440
the first place. And so you're going to end up getting,

262
00:16:58.799 --> 00:17:01.600
like you said, hallucination airs is going to shoot out

263
00:17:02.120 --> 00:17:05.759
going down a different direction because of some some stock

264
00:17:05.839 --> 00:17:08.160
points you have in there that's not even supposed to

265
00:17:08.200 --> 00:17:14.000
be in the flow. It's it's just it's just amazing,

266
00:17:14.359 --> 00:17:17.160
and it's you know, like you said, there's a lot

267
00:17:17.200 --> 00:17:21.039
of risks that are that are are out there, and

268
00:17:23.200 --> 00:17:26.000
I'm wanted to ask you this question. Our AI generated

269
00:17:26.039 --> 00:17:31.799
phishing emails, voice cloning deep fakes real concerns for businesses

270
00:17:31.839 --> 00:17:36.319
today or is that still more future focused? Should they

271
00:17:36.519 --> 00:17:41.119
be concerned about AI being able to do that? Because

272
00:17:41.799 --> 00:17:45.039
I see tapes and stuff of other people's voices and

273
00:17:45.079 --> 00:17:47.200
then you know, you can just steal them and run.

274
00:17:47.640 --> 00:17:51.039
So how's that working out?

275
00:17:53.599 --> 00:17:54.000
All right?

276
00:17:54.039 --> 00:17:55.640
So let me let me go ahead and comment on

277
00:17:55.680 --> 00:17:59.839
what you asked. So, yes, deep fake has really changed

278
00:17:59.839 --> 00:18:05.240
the landscape of attacks where they're actually imitating an actual,

279
00:18:05.279 --> 00:18:07.920
true person and this can be done.

280
00:18:08.119 --> 00:18:08.880
Now.

281
00:18:09.279 --> 00:18:13.319
The issue with that whole you know, immulating somebody and

282
00:18:14.240 --> 00:18:16.319
saying that this is like if they say it's me

283
00:18:16.599 --> 00:18:19.960
and then using my voice. While what I've done with

284
00:18:20.160 --> 00:18:23.000
this all these years that I've never used my voice.

285
00:18:22.759 --> 00:18:25.240
So nobody's ever be able to copy my voice.

286
00:18:25.680 --> 00:18:28.839
So I've always been security conscious and people are not

287
00:18:29.119 --> 00:18:31.960
about voice, and voice is a big deal. People can

288
00:18:32.279 --> 00:18:36.559
literally imitate your voice and then put up your picture

289
00:18:37.079 --> 00:18:39.279
and then put you out on the internet like you

290
00:18:39.359 --> 00:18:42.119
are this person and you're you know, and you're doing

291
00:18:42.200 --> 00:18:43.559
ABC whatever that is.

292
00:18:43.720 --> 00:18:45.279
You're you're you're you're.

293
00:18:45.160 --> 00:18:48.160
Asking for a password, you're saying Hey, I need to

294
00:18:48.160 --> 00:18:51.559
get into your network. It's an emergency. I got locked

295
00:18:51.559 --> 00:18:53.839
out and I need to take care of this. Uh

296
00:18:54.039 --> 00:18:56.559
this this this email from the from the boss or

297
00:18:56.720 --> 00:18:58.559
you know, they make up whatever story is they can

298
00:18:58.680 --> 00:19:02.400
and then you use behind you know that is unbeknownst

299
00:19:02.400 --> 00:19:04.200
to they use it. I think it's the true person,

300
00:19:04.599 --> 00:19:06.440
and so they go ahead and give that information.

301
00:19:06.839 --> 00:19:08.599
So that is still happening.

302
00:19:08.319 --> 00:19:11.519
That's still occurring, and all of this is it's user

303
00:19:11.680 --> 00:19:15.559
education really ultimately, it's the user needs to be educated

304
00:19:17.039 --> 00:19:20.119
on what the attack looks like. They need to have

305
00:19:20.160 --> 00:19:23.559
some sessions about that, you know. I mean there's so

306
00:19:23.640 --> 00:19:26.519
many like no before, there's so many applications out there

307
00:19:26.599 --> 00:19:30.319
that will go through all these videos to show you

308
00:19:30.359 --> 00:19:33.240
what's happening and what to be aware of. Right, So

309
00:19:34.319 --> 00:19:36.480
it's if there's definitely information out there, you just have

310
00:19:36.519 --> 00:19:39.119
to be aware not and not you know, take it

311
00:19:39.160 --> 00:19:40.880
for granted that that's the actual person.

312
00:19:41.400 --> 00:19:43.839
Yeah, you're back.

313
00:19:47.440 --> 00:19:50.440
Yeah, you know, I was just thinking about them, you know,

314
00:19:50.519 --> 00:19:53.680
people using voices and like you said, they just grab

315
00:19:53.759 --> 00:19:57.799
your picture and they you know, pretend that they're you're

316
00:19:57.960 --> 00:20:02.519
they're you. And as you were explaining that, you know,

317
00:20:02.640 --> 00:20:05.440
people need to get some training. That's why they need

318
00:20:05.480 --> 00:20:08.119
services like yours, because they just go out there and

319
00:20:08.160 --> 00:20:10.079
they think, oh, all I need to do is go

320
00:20:10.160 --> 00:20:14.079
get this little thing. So it's you know, checking this

321
00:20:14.200 --> 00:20:17.319
and checking that out. But there's some retraining that needs

322
00:20:17.359 --> 00:20:20.160
to happen. There's a lot of education needs to happen.

323
00:20:20.480 --> 00:20:23.480
And this is why everything, you know, these companies are

324
00:20:23.599 --> 00:20:27.200
failing because they've got people rushing in there saying, hey,

325
00:20:27.359 --> 00:20:30.680
let's go in there to help you use AI better.

326
00:20:31.799 --> 00:20:34.680
But do you understand everything that you need to know

327
00:20:34.839 --> 00:20:38.440
about this AI and the risk they are there. I

328
00:20:38.519 --> 00:20:41.400
don't think they're having those discussions. They just went to

329
00:20:41.720 --> 00:20:45.240
a class to learn how to set up algorithms in

330
00:20:45.279 --> 00:20:48.400
a six week you know program, and now you're out

331
00:20:48.400 --> 00:20:51.920
there marketing yourself is less like you're an expert in AI.

332
00:20:52.319 --> 00:20:54.960
And I'm really concerned about that because I'm seeing, like

333
00:20:55.000 --> 00:20:59.160
I said, companies are actually buying those services and then

334
00:20:59.200 --> 00:21:03.559
they're getting burned because it really requires a lot more

335
00:21:03.759 --> 00:21:06.920
knowledge in that. So you need to rely on the experts.

336
00:21:07.240 --> 00:21:13.640
Go ahead, yeah, go ahead.

337
00:21:19.720 --> 00:21:21.519
Of fit making is what it's all about.

338
00:21:22.119 --> 00:21:26.440
And so we are being we are being reckless when

339
00:21:26.480 --> 00:21:32.079
it comes to rolling out AI because it can. It

340
00:21:32.119 --> 00:21:37.400
can improve profit margins in the short term by letting

341
00:21:37.599 --> 00:21:41.640
you know, laying people off, letting people go, and speeding

342
00:21:41.720 --> 00:21:46.599
up processes. Granted if these processes are good processes, so

343
00:21:46.640 --> 00:21:48.079
there are some good things that can happen.

344
00:21:49.200 --> 00:21:50.359
But the vendors that.

345
00:21:50.359 --> 00:22:00.000
Are selling the AI tooling, yeah open ais chat cheep,

346
00:22:00.519 --> 00:22:05.440
and you know all those guys are out to make money.

347
00:22:06.559 --> 00:22:12.079
They are not providing guardrails. They are not provided I

348
00:22:12.160 --> 00:22:14.240
don't guardrails period.

349
00:22:14.759 --> 00:22:15.319
They don't care.

350
00:22:15.400 --> 00:22:19.880
They think we should develop our guardrails. That's all well

351
00:22:19.920 --> 00:22:22.920
and good good. So we are in the middle ground

352
00:22:23.000 --> 00:22:27.279
right now where there are no guardrails and there is

353
00:22:27.319 --> 00:22:37.839
a massively intelligent amount of processing available freely, not freely,

354
00:22:37.880 --> 00:22:41.400
but available to almost anybody to do almost anything. So

355
00:22:41.480 --> 00:22:45.200
these are very dangerous times. And I know that might sound,

356
00:22:45.240 --> 00:22:47.720
you know, pretty dark and sky is falling, but we

357
00:22:47.839 --> 00:22:54.759
have to be realistic right right now because, as I

358
00:22:54.839 --> 00:22:58.640
was mentioning before I got dropped, connections, zero trust is

359
00:22:58.680 --> 00:23:03.599
an important facet. Now you have to verify virtually everything

360
00:23:03.920 --> 00:23:07.680
that comes across you, at least digitally, right, and if

361
00:23:07.720 --> 00:23:10.799
that's a phone call from your supposed grandson, saying he's

362
00:23:10.839 --> 00:23:15.000
in trouble. You need to say if word to know

363
00:23:15.119 --> 00:23:17.799
that it's him. These are the times we live in.

364
00:23:18.319 --> 00:23:21.200
I wish it wasn't so, but we can be reckless

365
00:23:21.240 --> 00:23:23.759
and just say let Ai do it. Ai is super

366
00:23:23.799 --> 00:23:28.119
fast and smart. It's a machine that much theoretically smart,

367
00:23:29.079 --> 00:23:40.079
harder the next time it does same thing. But there

368
00:23:40.079 --> 00:23:45.599
are no okay, there's nothing there. So these are really

369
00:23:45.720 --> 00:23:49.720
dangerouous times, and I encourage the listeners out there to

370
00:23:49.799 --> 00:23:52.519
take it seriously. No, I don't want to. I want

371
00:23:52.559 --> 00:23:54.720
this to be successful. I don't want to be the

372
00:23:54.839 --> 00:23:56.519
guy who's saying.

373
00:23:57.839 --> 00:24:00.920
It's bad. As bad as it's bad.

374
00:24:01.240 --> 00:24:07.279
But I've lived my entire career, which is decades in cybersecurity,

375
00:24:07.759 --> 00:24:12.039
watching out for bad actors, bad situations, and we're a

376
00:24:12.079 --> 00:24:15.640
place right now where it has never ever been easier

377
00:24:15.680 --> 00:24:20.880
for bad actors to take advance advantage of good people,

378
00:24:20.960 --> 00:24:25.400
good organizations. And I think you're fully not but it

379
00:24:25.519 --> 00:24:26.440
is that way right now.

380
00:24:27.160 --> 00:24:33.599
Yeah, and yeah, I agree. It is one of those

381
00:24:33.640 --> 00:24:37.680
things where you know and believe me. I started with

382
00:24:38.119 --> 00:24:42.119
the first personal computer. I love technology, I love the

383
00:24:42.160 --> 00:24:45.880
growth of technology, and I'm excited to see this, you know,

384
00:24:46.559 --> 00:24:50.519
new evolution that has occurred with the AI. It's like,

385
00:24:50.599 --> 00:24:53.480
oh wow, this is so cool, especially when you started

386
00:24:53.519 --> 00:24:55.920
out with four tran and the punch cars and now

387
00:24:56.079 --> 00:25:04.160
you have these I'm aging myself Angela, and now you

388
00:25:04.240 --> 00:25:08.680
have these tools that can just immediately, within seconds just

389
00:25:08.799 --> 00:25:11.240
give you, you know, all the data that you've been

390
00:25:11.559 --> 00:25:13.839
searching for in the past and things like that. But

391
00:25:15.119 --> 00:25:17.799
just like as we started there, there were certain steps

392
00:25:17.839 --> 00:25:20.799
we had to go through to actually learn how to

393
00:25:20.880 --> 00:25:24.240
make sure the data that we were creating, the report

394
00:25:24.319 --> 00:25:27.680
that we were trying to develop, or whatever we were doing,

395
00:25:27.960 --> 00:25:30.720
we had to double check and double check that it

396
00:25:30.960 --> 00:25:34.440
actually followed all of the language that we had to

397
00:25:34.480 --> 00:25:36.799
write out and everything, and all the steps and all

398
00:25:36.839 --> 00:25:39.440
the loops and everything like that. We had to make

399
00:25:39.480 --> 00:25:43.599
sure that all of them actually were done because if

400
00:25:43.640 --> 00:25:47.440
one loop wasn't followed, the information was going to be wrong.

401
00:25:47.759 --> 00:25:51.240
It's no different. You have to do this with this

402
00:25:51.400 --> 00:25:54.880
tool so that it's doing it. You know also, but

403
00:25:54.920 --> 00:25:56.599
everybody's want to be the first. I want to be

404
00:25:56.599 --> 00:25:58.720
the first to do it. I want to be the first. Well,

405
00:25:58.759 --> 00:26:01.160
you're going to be the first to get burn because

406
00:26:02.119 --> 00:26:06.039
you don't have the protection and the knowledge, and you're

407
00:26:06.079 --> 00:26:10.440
setting your company up, like Doug was saying, to lose

408
00:26:10.559 --> 00:26:14.839
your trust factor. And that's what where your profity is.

409
00:26:14.839 --> 00:26:18.119
It's not in how many people are working or not working,

410
00:26:18.480 --> 00:26:22.240
it's how many people trust your company and trust your

411
00:26:22.279 --> 00:26:26.799
product and trust your services. You know what I'm saying

412
00:26:26.839 --> 00:26:31.680
exactly exactly what are some of the first cybersecurity steps

413
00:26:31.839 --> 00:26:35.000
every business should be taken. You know, even if they

414
00:26:35.039 --> 00:26:40.079
have a limited budget, what should they be doing? Tenevieed

415
00:26:41.039 --> 00:26:44.640
So number one is what I've noticed a lot in

416
00:26:44.680 --> 00:26:45.960
the industry is.

417
00:26:46.480 --> 00:26:48.599
The actual access to systems.

418
00:26:49.240 --> 00:26:52.200
So that is the very first place believe or not

419
00:26:52.279 --> 00:26:53.359
that we shouldn't be looking at.

420
00:26:53.400 --> 00:26:53.799
I am.

421
00:26:53.920 --> 00:26:56.880
It's called right, So identity access right.

422
00:26:57.920 --> 00:27:01.519
We need to be sure that whoever is getting access

423
00:27:01.680 --> 00:27:06.359
to the server, to the share drives to other people's computers,

424
00:27:06.680 --> 00:27:09.519
that they have the type of access that they need,

425
00:27:09.720 --> 00:27:13.799
and we call that as needed access right, So they

426
00:27:13.839 --> 00:27:18.200
don't need to know everything about everybody or all the files.

427
00:27:18.200 --> 00:27:21.720
So like there should be a hierarchy that should be followed.

428
00:27:21.720 --> 00:27:24.720
There should be like the accounting department, the HR department,

429
00:27:24.759 --> 00:27:29.079
the IT department. Everybody has their role and the responsibilities

430
00:27:29.160 --> 00:27:32.319
and the proper access, and we shouldn't have like five

431
00:27:32.400 --> 00:27:35.400
or six domain admins on the domain controller.

432
00:27:35.920 --> 00:27:37.279
All that happens when.

433
00:27:37.160 --> 00:27:40.519
We do that is we allow the hacker then has

434
00:27:40.720 --> 00:27:44.599
way more opportunities to Now, if he can't get into

435
00:27:44.599 --> 00:27:46.480
that account, he'll try the next account. He tried, the

436
00:27:46.519 --> 00:27:49.480
next account, he'll keep trying until he gets in. Right

437
00:27:49.960 --> 00:27:54.079
once again, it's over game over, right, And they have

438
00:27:54.240 --> 00:27:57.440
tools that do that very fast and very easy. So

439
00:27:57.559 --> 00:28:01.000
that's the first place is my thinking. And then we

440
00:28:01.039 --> 00:28:03.759
need to make sure that we have the proper set

441
00:28:03.839 --> 00:28:07.759
up in the environment architecturally, meaning we need to have

442
00:28:08.279 --> 00:28:10.680
set up a flat network. We need to have a

443
00:28:10.759 --> 00:28:15.880
hierarchy network where everything is segregated, where zero trust, everything's segregated,

444
00:28:16.480 --> 00:28:20.640
and we have you know that for example, you're in finance,

445
00:28:21.000 --> 00:28:23.519
you should be the only department that has access to

446
00:28:23.559 --> 00:28:27.119
that and maybe HR, but nobody else in the ID department,

447
00:28:27.160 --> 00:28:30.559
any other department should have access to your information because

448
00:28:30.599 --> 00:28:34.640
that's really confidential information that you don't want anybody else

449
00:28:34.680 --> 00:28:38.160
to know how much money you're making. Right, So it

450
00:28:38.200 --> 00:28:40.359
all oils down to that, And like you said before,

451
00:28:40.839 --> 00:28:43.240
we have to have the proper processes in places we

452
00:28:43.319 --> 00:28:46.200
have to have the proper policies in places, right, Doug,

453
00:28:46.400 --> 00:28:49.000
I mean, it just goes. These are the real core

454
00:28:49.799 --> 00:28:53.240
of meat, meat of the actual business. Before you go

455
00:28:53.279 --> 00:28:55.720
off and start implementing a bunch.

456
00:28:55.559 --> 00:28:59.680
Of tools, they need to get foundation set up first, right,

457
00:29:00.079 --> 00:29:03.839
you know. Again, it has to be and companies don't

458
00:29:03.839 --> 00:29:06.559
do that, and if they do, then it gets lost.

459
00:29:06.839 --> 00:29:07.160
What it.

460
00:29:07.480 --> 00:29:10.519
Departments start giving you more access than you need to information,

461
00:29:11.000 --> 00:29:12.720
and then before you know it, that person goes to

462
00:29:12.759 --> 00:29:15.000
another department and they still have the same access to

463
00:29:15.039 --> 00:29:16.960
where they were in the first department, but now they

464
00:29:17.000 --> 00:29:20.240
have bad access plus more access. So it's just a

465
00:29:20.519 --> 00:29:23.880
really bad cycle that happens.

466
00:29:24.200 --> 00:29:30.480
Yeah, oh go ahead, Angel, I'm sorry, you know.

467
00:29:30.680 --> 00:29:32.400
And and but wait, there's more.

468
00:29:36.680 --> 00:29:39.559
There's more. Yes, Tell has done.

469
00:29:40.160 --> 00:29:42.160
Yeah, yeah, I know there's a little delay, and I

470
00:29:42.200 --> 00:29:46.039
apologize for our listeners. We're trying to make do here

471
00:29:46.200 --> 00:29:50.519
strewn about the globe. But Genevieve, what you said is

472
00:29:50.640 --> 00:29:54.240
absolutely spot on. But there's there's more even to that,

473
00:29:54.400 --> 00:29:58.000
because yes, I've been working in identity and access management

474
00:29:58.039 --> 00:30:00.400
like you have for many, many, many years decades now,

475
00:30:00.480 --> 00:30:05.279
and most organizations of large size, medium size, and some

476
00:30:05.440 --> 00:30:10.480
small size don't do identity management very well. They do

477
00:30:10.759 --> 00:30:17.279
more coarse grained authorization. And you know, the news media

478
00:30:17.480 --> 00:30:22.200
is littered with stories of breaches where sensitive information was

479
00:30:22.200 --> 00:30:27.440
stolen and misused right by hackers, by people impersonating people

480
00:30:27.480 --> 00:30:31.599
in the organization and things like that. But why does

481
00:30:31.640 --> 00:30:35.640
it get worse because of AI agents? We need to

482
00:30:35.680 --> 00:30:38.440
think of AI agents as people too, all right, they're

483
00:30:38.480 --> 00:30:40.960
non carbon based life forms because they are life forms.

484
00:30:41.000 --> 00:30:45.599
They're learning, they're evolving, they're drifting, they're hallucinating, they're doing

485
00:30:45.640 --> 00:30:49.720
all kinds of things, right, and if we don't really

486
00:30:49.759 --> 00:30:52.319
lock them down, who put this Like things like who

487
00:30:52.319 --> 00:30:56.200
put this AI agent in? What is its intended purpose?

488
00:30:56.759 --> 00:30:59.319
How long is it supposed to live? What data should

489
00:30:59.319 --> 00:31:02.039
it have access to and only that data? What should

490
00:31:02.039 --> 00:31:04.400
it be able to do with that data? All of

491
00:31:04.400 --> 00:31:07.359
these kinds of things apply to agents, but they're not

492
00:31:07.400 --> 00:31:09.759
being rolled out that way people who are saying, put

493
00:31:09.759 --> 00:31:12.079
that agent out there, give it access to all our

494
00:31:12.160 --> 00:31:15.720
data and see what it tells us when I ask

495
00:31:15.759 --> 00:31:19.000
it a question that, my friends, is the worst governance

496
00:31:19.200 --> 00:31:22.559
you could ever imagine in terms of making your prized

497
00:31:22.599 --> 00:31:27.759
intellectual property, your sensitive customer data, every financial information hr

498
00:31:27.880 --> 00:31:31.440
information at huge risk because the agent. Listen, the agent

499
00:31:31.480 --> 00:31:34.000
itself isn't a bad guy. The agent just doesn't have

500
00:31:34.039 --> 00:31:37.279
any constraints. Nobody's watching it or telling it exactly what

501
00:31:37.400 --> 00:31:39.640
to do. Is saying, find me the answer to this,

502
00:31:39.680 --> 00:31:41.480
and here's access to all our data to find it.

503
00:31:41.680 --> 00:31:44.240
That's how stuff is being rolled out right now. That

504
00:31:44.720 --> 00:31:45.720
is catastrophic.

505
00:31:46.319 --> 00:31:52.240
Yes, yeah, I've been seeing this. And as you were mentioning,

506
00:31:52.799 --> 00:31:57.079
both of you were mentioning the fact that there's no restrictions.

507
00:31:57.440 --> 00:32:03.279
There's no restrictions on em on the carbon people, the

508
00:32:03.279 --> 00:32:06.759
people that are there, You're right, they have access to

509
00:32:06.920 --> 00:32:12.119
information that it's not even part of their job responsibilities.

510
00:32:12.119 --> 00:32:14.039
What are they doing being able to see this? I mean,

511
00:32:14.200 --> 00:32:16.240
and I'm going there all the time. I have to

512
00:32:16.279 --> 00:32:19.480
always be the bad guy. I don't care. I wear

513
00:32:19.519 --> 00:32:22.400
my hat. I like it. I'm going to tell you

514
00:32:22.480 --> 00:32:25.480
what are you doing? Why are they seeing this information?

515
00:32:25.559 --> 00:32:28.599
What I have to do with their responsibilities? Nothing? And

516
00:32:28.960 --> 00:32:31.200
get this off of here, like right now.

517
00:32:31.079 --> 00:32:31.440
Me too.

518
00:32:31.880 --> 00:32:34.680
And when these people leave, you ought to have an automated,

519
00:32:35.000 --> 00:32:38.279
automated system that says these people have moved moved from

520
00:32:38.279 --> 00:32:43.640
this department and it wipes out their access to the

521
00:32:43.680 --> 00:32:46.960
tools that they were in automatically so that you don't

522
00:32:47.519 --> 00:32:49.720
have to worry. Oh I forgot, I didn't know. They

523
00:32:49.799 --> 00:32:54.359
move nobody communicated. As soon as they get new access,

524
00:32:54.440 --> 00:32:56.640
their other is cut off. But you can't let them

525
00:32:56.720 --> 00:32:59.319
just have access to everything. And this is I mean,

526
00:32:59.480 --> 00:33:02.440
we just walking around just mis well, just open all

527
00:33:02.480 --> 00:33:04.839
the windows and all the doors and say, come on in,

528
00:33:05.319 --> 00:33:09.640
because you're sending out invitations. That's what they're doing, sending

529
00:33:09.640 --> 00:33:12.960
out invitations. Come on in, hack me take everything I

530
00:33:13.000 --> 00:33:18.440
got because I I like being open minded. I don't

531
00:33:18.440 --> 00:33:26.559
know what that is. So that go ahead, go ahead.

532
00:33:30.640 --> 00:33:33.799
That is the state of AI today today. And I

533
00:33:33.799 --> 00:33:36.000
hate to break it to the listeners, but that is

534
00:33:36.200 --> 00:33:39.240
the state of AI today. It's not a good state.

535
00:33:39.599 --> 00:33:44.680
That's why I said earlier. Governance is very very much needed.

536
00:33:45.319 --> 00:33:48.960
And until such governance that we can trust is put

537
00:33:49.000 --> 00:33:55.960
in place, I implore your listeners to, I don't say,

538
00:33:56.599 --> 00:33:59.759
be very wary of trusting anything that's anything that's coming

539
00:33:59.759 --> 00:34:03.400
from AI or from a vendor or an apper. And

540
00:34:03.440 --> 00:34:06.640
if we're using AI, AI is doing this, AI is

541
00:34:06.680 --> 00:34:09.239
doing whatever the heck it wants with your data. You

542
00:34:09.320 --> 00:34:11.559
need to know that there is no governance structure in

543
00:34:11.599 --> 00:34:14.079
place right now. There are no regulations. There is no

544
00:34:14.280 --> 00:34:17.599
government in the United States that wants regulations at the

545
00:34:17.599 --> 00:34:19.760
present time. This is what we live in and you

546
00:34:19.880 --> 00:34:21.079
have to be careful.

547
00:34:22.400 --> 00:34:27.599
And that's scary that there's no You're right, there are

548
00:34:27.639 --> 00:34:31.760
no regulations. There's no laws as to what AI should

549
00:34:31.840 --> 00:34:34.039
be allowed to do and what it should not be

550
00:34:34.159 --> 00:34:36.400
allowed to do. And there are some things it should

551
00:34:36.400 --> 00:34:38.920
not be allowed to do that you know, when it's

552
00:34:38.920 --> 00:34:42.920
starting to get into people's personal privacy and different things

553
00:34:42.960 --> 00:34:46.239
like that. It shouldn't even have capability to do some

554
00:34:46.360 --> 00:34:48.880
of the things that it has right now, like it's

555
00:34:48.960 --> 00:34:52.719
just an open field, and why someone would not want

556
00:34:52.760 --> 00:34:57.440
to put guard rails. Then that makes me concerned as

557
00:34:57.480 --> 00:35:00.199
to what the intent is. What do you need that

558
00:35:00.320 --> 00:35:03.199
information for? It's always about what do you need that

559
00:35:03.320 --> 00:35:05.719
data for? For What purpose are you using it? If

560
00:35:05.719 --> 00:35:09.719
you're not using it for something to benefit, then you

561
00:35:09.760 --> 00:35:11.840
don't need to see it. You don't need access, and

562
00:35:11.920 --> 00:35:17.320
we need to have rules and guardrails, you know, and

563
00:35:17.719 --> 00:35:19.119
emphasize that enough.

564
00:35:25.400 --> 00:35:28.760
By the time you find out that your data has

565
00:35:28.840 --> 00:35:32.920
been misused by an AI agent, your data is already

566
00:35:32.960 --> 00:35:35.760
one hundred miles down the road never to return.

567
00:35:36.320 --> 00:35:42.360
Yeah, and then everybody is vulnerable because of it. You know,

568
00:35:42.480 --> 00:35:46.639
do some you know when we're talking about this, I

569
00:35:46.679 --> 00:35:51.719
don't want small companies to feel like they shouldn't be

570
00:35:51.800 --> 00:35:55.639
included in this conversation. Every company needs to be included,

571
00:35:55.679 --> 00:35:57.599
don't you think absolutely?

572
00:35:58.480 --> 00:36:01.320
Actually, you like to go after the small people because

573
00:36:01.320 --> 00:36:04.760
they're easy targets. People always think, oh, hackers only go

574
00:36:04.800 --> 00:36:06.480
after the big companies with all the money.

575
00:36:06.519 --> 00:36:09.199
Well guess what, Yes, that's not sophisticated.

576
00:36:09.920 --> 00:36:12.719
Yep, because because they know that a lot of people

577
00:36:12.719 --> 00:36:15.360
don't take care of their passwords, they don't take care

578
00:36:15.400 --> 00:36:17.559
of their loggings to all this. You know, no, not

579
00:36:17.599 --> 00:36:21.400
everybody uses a password manageer management database, and they should.

580
00:36:21.880 --> 00:36:23.519
They're free in the needs.

581
00:36:24.239 --> 00:36:26.320
And you don't want to be typing in your password.

582
00:36:26.360 --> 00:36:28.519
You want to let your password manager type of.

583
00:36:30.079 --> 00:36:33.599
Input if you will. So that's it.

584
00:36:35.440 --> 00:36:40.719
Yeah, So all of you new use new businesses, small businesses,

585
00:36:41.119 --> 00:36:46.079
you know, emerging businesses, growing businesses. You are vulnerable too,

586
00:36:46.679 --> 00:36:52.400
and you need to contact my guest to give you

587
00:36:52.440 --> 00:36:55.360
some advice of what you need to do to be protected.

588
00:36:55.440 --> 00:36:57.480
The last thing you want to do is to be,

589
00:36:58.119 --> 00:37:04.079
you know, a victim of one of these cyber security attacks,

590
00:37:05.639 --> 00:37:08.679
because then they can ruin you and whatever your dream

591
00:37:08.880 --> 00:37:13.280
was is now gone and your reputation too. So it's

592
00:37:13.320 --> 00:37:17.880
for everyone, everyone that has any information that they are

593
00:37:18.039 --> 00:37:22.719
utilizing in a business. Whatsoever, Go ahead, k.

594
00:37:26.320 --> 00:37:31.480
May I say yes, go ahead. These issues are not insurmountable.

595
00:37:31.559 --> 00:37:33.920
I want to stress that. I want to stress for

596
00:37:33.960 --> 00:37:36.840
the readers that these issues are not insurmountable, but they

597
00:37:36.920 --> 00:37:40.199
require governance. They require us human beings to have our

598
00:37:40.239 --> 00:37:43.599
eyes wide open. The people like the information officer, the

599
00:37:43.800 --> 00:37:48.320
chief Information Security Officer, the CISO, everyone responsible for data

600
00:37:48.320 --> 00:37:52.840
protection and financial information, intellectual property, etc. Need to know

601
00:37:53.199 --> 00:37:57.800
that governance is possible, but governance is hard, not impossible.

602
00:38:00.519 --> 00:38:02.880
We, however, Genevieve and I, we have.

603
00:38:03.719 --> 00:38:06.920
Very well thought out checklists of what you need to do,

604
00:38:07.159 --> 00:38:09.639
and we can scratch the surface in this discussion just

605
00:38:09.639 --> 00:38:14.119
by saying, discover what agents are running in your enterprise.

606
00:38:14.280 --> 00:38:18.320
Now make sure that every agent is registered in your

607
00:38:18.400 --> 00:38:22.360
identity and access management system. Now make sure that every

608
00:38:22.400 --> 00:38:26.199
agent has a human owner responsible for the actions of

609
00:38:26.239 --> 00:38:30.360
that agent. Now, these things are actually simple, right, They

610
00:38:30.400 --> 00:38:32.719
don't require a new technology or anything, but they require

611
00:38:32.719 --> 00:38:34.840
everybody to get up and say we're going to do this.

612
00:38:35.559 --> 00:38:40.800
Then you can start to put in monitoring guardrails around

613
00:38:40.840 --> 00:38:43.519
what an AI agent can do because now it's registered

614
00:38:43.519 --> 00:38:47.199
in IAM, so we'll have access control privileges, meaning, like

615
00:38:48.280 --> 00:38:50.559
Genevieve said, things that it can do and things that

616
00:38:50.840 --> 00:38:55.239
explicitly cannot do or information it cannot access. And then

617
00:38:55.280 --> 00:38:58.320
we have to have ongoing monitoring cybersecurity tools like we

618
00:38:58.400 --> 00:39:01.239
have to make sure that those rules are being followed.

619
00:39:01.559 --> 00:39:07.000
So all we're saying is that AI record why there's rules,

620
00:39:07.480 --> 00:39:11.559
just like the rest of us, and that is not insurmountable.

621
00:39:11.960 --> 00:39:16.079
But the real quick roll this out lay people off,

622
00:39:16.159 --> 00:39:20.280
make more money. Mindset needs to be reeled back rapidly

623
00:39:20.679 --> 00:39:23.320
so that we can focus more on roll this out,

624
00:39:23.360 --> 00:39:26.679
but only when it's properly governed to make sure our

625
00:39:26.719 --> 00:39:29.239
company and our people is not at risk. We can

626
00:39:29.280 --> 00:39:31.599
do that today, but you have to know what you're

627
00:39:31.639 --> 00:39:32.480
doing and how to do it.

628
00:39:32.519 --> 00:39:37.320
We can help you, yes, absolutely, you know. I think

629
00:39:37.360 --> 00:39:40.519
what because like I said, it makes these cute little

630
00:39:41.000 --> 00:39:45.719
you know, videos and different things like that, people are overconfident.

631
00:39:45.760 --> 00:39:49.239
In fact, they've created this false sense of security with AI,

632
00:39:49.440 --> 00:39:53.519
like AI is going to save the day. And you

633
00:39:53.559 --> 00:39:58.079
know they think, as we can replace human judgment, and

634
00:39:58.159 --> 00:40:00.760
like I said, they some have found the hard way.

635
00:40:00.880 --> 00:40:05.719
I mean, I'm not talking about small companies who are vulnerable.

636
00:40:06.039 --> 00:40:09.360
I'm talking about huge companies who did not think they

637
00:40:09.400 --> 00:40:17.400
were vulnerable and have lost their trust because they just did.

638
00:40:17.559 --> 00:40:21.920
They assumed that AI was the be all of everything,

639
00:40:22.440 --> 00:40:27.960
and it does require some guardrails, some restrictions, someone who

640
00:40:28.039 --> 00:40:32.679
really understands how it's working in tandem with all other

641
00:40:32.840 --> 00:40:37.320
systems that it's connected to and potential ones that are

642
00:40:37.360 --> 00:40:40.679
trying to get in. You need experts in there looking

643
00:40:40.719 --> 00:40:43.480
at what's going on. Not you know, this is not

644
00:40:43.599 --> 00:40:47.119
a time to be, you know, out there like you're

645
00:40:47.159 --> 00:40:50.800
playing with your toys. This is serious business here because

646
00:40:50.840 --> 00:40:57.760
you have made yourself vulnerable and you need to understand

647
00:40:58.239 --> 00:41:01.800
you probably have more risk now than you've ever had

648
00:41:02.760 --> 00:41:05.320
because of some of the decisions you have just made.

649
00:41:06.039 --> 00:41:12.320
What scares you more? AI powered at tax or overconfident?

650
00:41:12.920 --> 00:41:17.039
These over confident companies that think AI will save them.

651
00:41:17.760 --> 00:41:21.079
Which one do you use? Which one is worse for you?

652
00:41:22.519 --> 00:41:22.840
All right?

653
00:41:22.920 --> 00:41:25.280
Dad, Okay, I'll go first.

654
00:41:25.320 --> 00:41:28.559
They are two sides of the exact same coin, unfortunately, right,

655
00:41:28.800 --> 00:41:32.360
because if you have an organization that does not govern

656
00:41:32.599 --> 00:41:37.199
AI properly, then it cannot determine if AI is doing

657
00:41:37.239 --> 00:41:40.920
something wrong, meaning that if an AI agent is coming

658
00:41:40.920 --> 00:41:44.760
in from the outside, they don't have the appropriate cybersecurity

659
00:41:44.760 --> 00:41:47.920
tools to know that it's being duped, it's being fooled,

660
00:41:47.960 --> 00:41:51.199
it's not really the enterprise isn't really talking to who

661
00:41:51.280 --> 00:41:52.679
or what it thinks it's talking to.

662
00:41:53.119 --> 00:41:55.480
Right, So these are two sides of the same coin.

663
00:41:55.599 --> 00:41:56.639
Now where would I begin?

664
00:41:57.719 --> 00:42:00.639
Oh boy, I don't think you can say start on

665
00:42:01.199 --> 00:42:05.480
the cybersecurity tooling using AI. That's a good place to

666
00:42:05.519 --> 00:42:08.159
start because remember what I said at the beginning. You

667
00:42:08.199 --> 00:42:11.719
effectively get a million little guys doing the cybersecurity work

668
00:42:11.800 --> 00:42:15.280
for you, and that's really helpful because you can see

669
00:42:15.800 --> 00:42:19.159
and monitor and evaluate a lot more, a lot faster.

670
00:42:19.480 --> 00:42:21.199
But you still need to come to a human being

671
00:42:21.360 --> 00:42:24.440
to say yes, no or otherwise. Right, that's good. So

672
00:42:24.519 --> 00:42:29.239
get your cybersecurity tools AI empowered, but don't trust AI

673
00:42:29.400 --> 00:42:31.599
out of the box. Even in your tools, make sure

674
00:42:31.639 --> 00:42:33.559
you know what the boundaries are for that tools to do,

675
00:42:34.280 --> 00:42:36.400
know your vendor, and know the tool that you're using

676
00:42:37.000 --> 00:42:41.320
not rocket science easily doable on the internal side as

677
00:42:41.320 --> 00:42:45.880
far as using AI to do your business processing and automation,

678
00:42:46.239 --> 00:42:50.000
and let's just say cost reduction, which is a major

679
00:42:50.039 --> 00:42:53.039
reason for it. Do the things that I just said,

680
00:42:53.079 --> 00:42:56.039
I won't belong winded and say it again. Register every agent,

681
00:42:56.199 --> 00:42:58.960
have an owner for every agent, have every agent with

682
00:42:59.039 --> 00:43:02.719
its access control privileges clearly defined and monitored, and then

683
00:43:02.760 --> 00:43:07.599
you can make good, good progress using AI internally. So

684
00:43:07.679 --> 00:43:11.760
in both cases, internally and from a tooling standpoint, you

685
00:43:11.840 --> 00:43:14.039
need to know what your AI is doing. You need

686
00:43:14.119 --> 00:43:16.320
to have ownership over it, you need to govern it

687
00:43:16.400 --> 00:43:19.199
clearly and strongly, and then you should be able to

688
00:43:19.199 --> 00:43:23.320
reap the benefits on both sides.

689
00:43:23.400 --> 00:43:26.320
But if you don't, you open yourself up to huge risk.

690
00:43:26.480 --> 00:43:30.239
Because AI is a being that is learning and will

691
00:43:30.480 --> 00:43:34.079
drift and will hallucinate. It goes without saying it always

692
00:43:34.079 --> 00:43:36.480
will because it doesn't have boundaries around it to keep

693
00:43:36.519 --> 00:43:38.519
it from doing. So that's how it is out of

694
00:43:38.559 --> 00:43:42.280
the box by on purpose. So it's incumbent on us

695
00:43:42.559 --> 00:43:44.280
to make sure the godrails are put in place.

696
00:43:45.079 --> 00:43:45.679
End of story.

697
00:43:47.960 --> 00:43:51.760
And I would also probably get some guidance as to

698
00:43:52.039 --> 00:43:57.559
let information the AI who really needs to be, you know,

699
00:43:57.960 --> 00:44:03.079
gathering information and learning from so you can limit it.

700
00:44:03.119 --> 00:44:05.480
If you've got one that's going to be handling all

701
00:44:05.519 --> 00:44:09.079
of the financial processes, it doesn't need to be in HR.

702
00:44:09.599 --> 00:44:13.400
If it's doing HR processes and getting information, it doesn't

703
00:44:13.440 --> 00:44:17.880
need to be in sales and marketing. Now I might

704
00:44:17.960 --> 00:44:20.440
be wrong, but I think segregating it, just like we're

705
00:44:20.480 --> 00:44:25.920
segregating your permissions, would really probably help out a lot

706
00:44:25.960 --> 00:44:29.639
instead of just having you know, maybe one AI tool

707
00:44:29.679 --> 00:44:34.159
that's just running through everything and becoming confused, especially seeing

708
00:44:35.760 --> 00:44:38.199
I think my last conversation I have with you is

709
00:44:38.280 --> 00:44:40.960
you don't like to do your SOPs. You don't have

710
00:44:41.000 --> 00:44:44.159
any policies, you don't follow any rules, you don't have

711
00:44:44.199 --> 00:44:48.840
any regulations, you don't even learn what the regulations are.

712
00:44:48.960 --> 00:44:53.239
You don't know what the guys are going. Then you're, yeah,

713
00:44:53.320 --> 00:44:57.599
you're just about to walk into a real mess. But yeah,

714
00:44:58.079 --> 00:45:01.679
I think you know everything that they're saying is really really,

715
00:45:01.760 --> 00:45:07.480
really really important. I'd like you guys that if you

716
00:45:07.519 --> 00:45:12.880
would share your contact information to our audience, because we're

717
00:45:12.880 --> 00:45:16.840
on radio and TV also, so if you could share

718
00:45:16.880 --> 00:45:20.639
your audience your contact information, that would be great.

719
00:45:22.639 --> 00:45:23.960
How would you like us to do that?

720
00:45:26.719 --> 00:45:30.920
If you can share your email and your website and

721
00:45:31.039 --> 00:45:33.400
if you have a contact phone number also.

722
00:45:34.159 --> 00:45:36.480
Where but where would you like me to put that information?

723
00:45:36.840 --> 00:45:38.800
Is it in the chat or do you want me

724
00:45:38.880 --> 00:45:39.840
to send that separately?

725
00:45:40.880 --> 00:45:43.239
If you could just say it right now? Because we all.

726
00:45:45.360 --> 00:45:46.639
Okay, so you can you.

727
00:45:46.559 --> 00:45:50.840
Can reach us at Intelligent Forensics X dot com. Uh

728
00:45:51.000 --> 00:45:56.239
of course www. And that's our website. On our website, email,

729
00:45:56.760 --> 00:45:59.599
our email address, and a phone number where we could

730
00:45:59.599 --> 00:46:02.719
be reached. So once you go to that site, the

731
00:46:02.760 --> 00:46:06.559
information is there, but our email is cyber at Intelligent

732
00:46:06.679 --> 00:46:11.599
Forensics x dot com. And then, like I said, all

733
00:46:11.639 --> 00:46:14.199
the other information is on there, but our phone number

734
00:46:14.280 --> 00:46:16.119
if you wanted to reach us by phone, it's two

735
00:46:16.119 --> 00:46:21.480
one three two five four five zero sixty six. That's

736
00:46:21.519 --> 00:46:24.239
our direct line, our business line, and we can be

737
00:46:24.280 --> 00:46:28.320
reached there. And Doug and I work very closely, so

738
00:46:28.360 --> 00:46:31.920
we're always available for you know, just to even even

739
00:46:31.960 --> 00:46:35.800
a consultation call zero costs, just to have an understanding

740
00:46:36.039 --> 00:46:38.880
of what you may be experiencing and where you need

741
00:46:38.920 --> 00:46:40.840
help and where you have gaps.

742
00:46:40.880 --> 00:46:42.280
We could do a gap analysis.

743
00:46:42.320 --> 00:46:44.559
We can come in and really have an understanding of

744
00:46:44.599 --> 00:46:47.920
your environment, your concerns. There's so many things that we

745
00:46:47.960 --> 00:46:52.079
can address once we have that conversation, So please feel

746
00:46:52.119 --> 00:46:54.800
free to reach out to Doug and I and.

747
00:46:56.400 --> 00:46:56.760
You can.

748
00:46:57.199 --> 00:47:00.440
I believe Doug, I don't know if you want to

749
00:47:00.480 --> 00:47:11.719
give additional information. If not, we could both be reached

750
00:47:11.719 --> 00:47:12.800
at the same contact.

751
00:47:13.400 --> 00:47:14.039
That's why.

752
00:47:14.480 --> 00:47:21.079
Okay, okay, Well this has been extremely informative. You know,

753
00:47:24.800 --> 00:47:30.039
if there's one takeaway that I would say today is

754
00:47:30.079 --> 00:47:36.400
that cybersecurity, it's not just about tools. It's about decisions,

755
00:47:36.679 --> 00:47:45.800
your culture, accountability, preparedness, you know, being aware, being informed.

756
00:47:46.519 --> 00:47:50.559
It's just a lot that you need to know, and

757
00:47:50.880 --> 00:47:54.719
it's really important for you to be aware of it

758
00:47:54.800 --> 00:47:58.679
and contact the experts. Thank you guys for being here today.

759
00:47:58.840 --> 00:48:03.880
It's really been one conversation and a lot for people

760
00:48:03.960 --> 00:48:09.079
to chew on today. And I'm just thankful that you

761
00:48:09.119 --> 00:48:12.800
were here to help our audience to be informed because

762
00:48:13.719 --> 00:48:15.840
there are a lot of mistakes that are happening. So

763
00:48:16.239 --> 00:48:18.760
I just want to thank our audience for being with

764
00:48:18.840 --> 00:48:21.920
us today to talk about this important topic that we

765
00:48:22.039 --> 00:48:27.599
have here. And as I always say, we're here every

766
00:48:27.679 --> 00:48:33.400
Monday at eleven am Eastern time eight am Pacific time,

767
00:48:33.880 --> 00:48:36.880
and we're here to help you be informed so you

768
00:48:37.000 --> 00:48:40.519
can be successful. Whether it's in your business, whether it's

769
00:48:40.559 --> 00:48:43.760
in your life. The one thing you always need to

770
00:48:43.800 --> 00:48:51.639
do is stay informed. I want to thank you for

771
00:48:51.760 --> 00:48:54.480
joining us on a sharp O book. We have been

772
00:48:54.480 --> 00:48:58.039
informed and energized to take the next steps. We have

773
00:48:58.159 --> 00:49:01.920
posted links to websites and videos to learn more on

774
00:49:02.000 --> 00:49:05.880
today's topic. Please join us again next week or another

775
00:49:06.000 --> 00:49:10.719
thought provoking conversation right here on k for HD radio

776
00:49:11.199 --> 00:49:15.239
and talk for TV. Listen to the podcast on all

777
00:49:15.599 --> 00:49:20.079
the podcast apps, and until next week, stay in Fard