WEBVTT

1
00:00:00.210 --> 00:00:01.200
In this lesson,

2
00:00:01.200 --> 00:00:04.410
we will learn about privacy risk considerations.

3
00:00:04.410 --> 00:00:06.600
Privacy risk encompasses the potential

4
00:00:06.600 --> 00:00:09.420
for unauthorized access, disclosure,

5
00:00:09.420 --> 00:00:12.750
or misuse of personal or sensitive information.

6
00:00:12.750 --> 00:00:16.860
Privacy risk considerations include the use of biometrics,

7
00:00:16.860 --> 00:00:19.920
data subject rights, and data sovereignty.

8
00:00:19.920 --> 00:00:23.100
Let's explore each of these privacy risk considerations

9
00:00:23.100 --> 00:00:24.330
in more detail.

10
00:00:24.330 --> 00:00:28.560
First, we have considerations specific to biometric data.

11
00:00:28.560 --> 00:00:32.490
Biometrics include technologies such as fingerprint scans,

12
00:00:32.490 --> 00:00:36.720
facial recognition, iris scans, and voice recognition.

13
00:00:36.720 --> 00:00:40.110
Biometric verification systems play an important role

14
00:00:40.110 --> 00:00:42.090
in identity verification,

15
00:00:42.090 --> 00:00:45.690
but also introduce privacy risks due to the highly sensitive

16
00:00:45.690 --> 00:00:49.230
and permanent nature of the biometric data they use.

17
00:00:49.230 --> 00:00:53.430
Additionally, not all biometric systems offer the same level

18
00:00:53.430 --> 00:00:55.650
of security or accuracy.

19
00:00:55.650 --> 00:00:57.870
In assessing biometric systems,

20
00:00:57.870 --> 00:01:00.390
two key metrics should be considered,

21
00:01:00.390 --> 00:01:02.160
the False Acceptance Rate,

22
00:01:02.160 --> 00:01:05.310
which measures how often unauthorized individuals

23
00:01:05.310 --> 00:01:08.520
are mistakenly identified as legitimate users,

24
00:01:08.520 --> 00:01:10.800
and the False Rejection Rate,

25
00:01:10.800 --> 00:01:13.110
which tracks how often legitimate users

26
00:01:13.110 --> 00:01:15.480
are wrongly denied access.

27
00:01:15.480 --> 00:01:18.150
Balancing these rates is essential.

28
00:01:18.150 --> 00:01:20.940
If the false acceptance rate is too high,

29
00:01:20.940 --> 00:01:22.530
security is at risk

30
00:01:22.530 --> 00:01:25.950
because unauthorized users may gain access.

31
00:01:25.950 --> 00:01:29.340
Conversely, if the false rejection rate is too high,

32
00:01:29.340 --> 00:01:32.640
legitimate users will be denied access.

33
00:01:32.640 --> 00:01:34.620
To fine-tune this balance,

34
00:01:34.620 --> 00:01:37.290
biometric sensitivity is adjusted.

35
00:01:37.290 --> 00:01:40.530
When sensitivity is adjusted on a biometric system,

36
00:01:40.530 --> 00:01:42.030
the false acceptance rate

37
00:01:42.030 --> 00:01:45.510
and false rejection rate behave in opposite ways.

38
00:01:45.510 --> 00:01:47.820
As sensitivity level increases,

39
00:01:47.820 --> 00:01:50.040
the system becomes more strict,

40
00:01:50.040 --> 00:01:53.100
causing the false rejection rate to rise.

41
00:01:53.100 --> 00:01:56.820
At the same time, the false acceptance rate decreases

42
00:01:56.820 --> 00:02:00.180
because the system is more selective in granting access.

43
00:02:00.180 --> 00:02:03.510
Conversely, when the sensitivity level is lowered,

44
00:02:03.510 --> 00:02:05.550
the system becomes more lenient,

45
00:02:05.550 --> 00:02:08.160
reducing the false rejection rate,

46
00:02:08.160 --> 00:02:11.100
but increasing the false acceptance rate.

47
00:02:11.100 --> 00:02:14.130
The Crossover Error Rate or CER

48
00:02:14.130 --> 00:02:16.440
is the point where the false acceptance rate

49
00:02:16.440 --> 00:02:18.960
and false rejection rates are equal.

50
00:02:18.960 --> 00:02:21.900
On a graph, this is where the false acceptance rate

51
00:02:21.900 --> 00:02:24.990
and false rejection rate curves intersect.

52
00:02:24.990 --> 00:02:27.180
System sensitivity should be adjusted

53
00:02:27.180 --> 00:02:30.480
to a point near the crossover error rate,

54
00:02:30.480 --> 00:02:32.430
where the false acceptance rate

55
00:02:32.430 --> 00:02:35.850
and the false rejection rate errors are minimized,

56
00:02:35.850 --> 00:02:39.570
ensuring both security and user convenience.

57
00:02:39.570 --> 00:02:42.270
Second, we have data subject rights.

58
00:02:42.270 --> 00:02:44.580
Data subject rights are legal protections

59
00:02:44.580 --> 00:02:46.560
that allow individuals control

60
00:02:46.560 --> 00:02:49.380
over their personal information data.

61
00:02:49.380 --> 00:02:52.800
Subject rights are outlined in various privacy laws,

62
00:02:52.800 --> 00:02:57.360
such as the General Data Protection Regulation or GDPR,

63
00:02:57.360 --> 00:03:01.200
the California Consumer Privacy Act, or CCPA

64
00:03:01.200 --> 00:03:03.420
and the New York SHIELD Act,

65
00:03:03.420 --> 00:03:05.400
where SHIELD stands for Stop Hacks

66
00:03:05.400 --> 00:03:08.280
and Improve Electronic Data Security.

67
00:03:08.280 --> 00:03:12.810
Under GDPR, individuals can access, correct, delete,

68
00:03:12.810 --> 00:03:15.180
and transfer their personal data,

69
00:03:15.180 --> 00:03:18.210
and they must be told how an organization intends

70
00:03:18.210 --> 00:03:21.840
to use their collected data before it is collected.

71
00:03:21.840 --> 00:03:26.160
Similarly, the CCPA provides California residents

72
00:03:26.160 --> 00:03:29.580
with the right to know what data is being collected.

73
00:03:29.580 --> 00:03:33.270
In addition, they could request that their data be deleted

74
00:03:33.270 --> 00:03:36.510
and opt out of any sale of their data.

75
00:03:36.510 --> 00:03:40.110
New York's SHIELD Act focuses on data security,

76
00:03:40.110 --> 00:03:43.080
ensuring organizations take reasonable steps

77
00:03:43.080 --> 00:03:44.880
to protect personal data

78
00:03:44.880 --> 00:03:48.120
and notify individuals in the event of a breach.

79
00:03:48.120 --> 00:03:51.750
These laws empower individuals who are referred to

80
00:03:51.750 --> 00:03:54.570
as the data subjects by providing them

81
00:03:54.570 --> 00:03:56.790
with the tools and rights to control

82
00:03:56.790 --> 00:03:59.760
how their personal information is collected,

83
00:03:59.760 --> 00:04:01.740
used, and shared.

84
00:04:01.740 --> 00:04:04.230
Third, we have data sovereignty.

85
00:04:04.230 --> 00:04:06.480
Data sovereignty refers to the principle

86
00:04:06.480 --> 00:04:09.750
that data is subject to the laws and regulations

87
00:04:09.750 --> 00:04:13.770
of the country in which it is collected and/or stored.

88
00:04:13.770 --> 00:04:16.590
This means that organization must follow

89
00:04:16.590 --> 00:04:18.660
the legal frameworks of both the country

90
00:04:18.660 --> 00:04:20.310
where the data originates

91
00:04:20.310 --> 00:04:23.310
and where it is eventually stored or processed.

92
00:04:23.310 --> 00:04:24.240
For instance,

93
00:04:24.240 --> 00:04:27.810
under the General Data Protection Regulation or GDPR,

94
00:04:27.810 --> 00:04:29.190
personal data collected

95
00:04:29.190 --> 00:04:32.340
from European Union citizens must comply

96
00:04:32.340 --> 00:04:34.230
with GDPR requirements,

97
00:04:34.230 --> 00:04:38.460
even if the data is stored outside of the European Union.

98
00:04:38.460 --> 00:04:42.450
If an American company collects data from European customers

99
00:04:42.450 --> 00:04:44.880
and stores it on servers in the US,

100
00:04:44.880 --> 00:04:49.234
that data is still subject to GDPR regulations.

101
00:04:49.234 --> 00:04:52.620
This adds complexity for multinational companies that need

102
00:04:52.620 --> 00:04:56.280
to ensure compliance with multiple legal standards.

103
00:04:56.280 --> 00:05:00.840
So data sovereignty ensures that data is handled responsibly

104
00:05:00.840 --> 00:05:03.870
and in accordance with local privacy laws

105
00:05:03.870 --> 00:05:07.410
regardless of where it is stored or processed.

106
00:05:07.410 --> 00:05:09.360
To better understand the complexity

107
00:05:09.360 --> 00:05:11.460
of privacy risk considerations,

108
00:05:11.460 --> 00:05:15.990
let's take a look at the 2019 BioStar 2 Data Breach.

109
00:05:15.990 --> 00:05:20.250
The BioStar 2 data breach raised concerns about biometrics,

110
00:05:20.250 --> 00:05:23.160
data subject rights, and data sovereignty.

111
00:05:23.160 --> 00:05:27.570
It exposed over 27.8 million biometric records,

112
00:05:27.570 --> 00:05:31.200
including fingerprints and facial recognition data.

113
00:05:31.200 --> 00:05:34.440
As biometric data is a permanent identifier,

114
00:05:34.440 --> 00:05:37.500
its exposure poses serious long-term risks.

115
00:05:37.500 --> 00:05:41.070
Under the General Data Protection Regulation or GDPR,

116
00:05:41.070 --> 00:05:43.502
this type of data is classified

117
00:05:43.502 --> 00:05:45.780
as a special category of data,

118
00:05:45.780 --> 00:05:47.580
requiring stronger safeguards

119
00:05:47.580 --> 00:05:50.190
to prevent unauthorized access.

120
00:05:50.190 --> 00:05:52.680
This breach also highlighted questions

121
00:05:52.680 --> 00:05:54.615
about whether Suprema,

122
00:05:54.615 --> 00:05:56.940
the company behind BioStar 2,

123
00:05:56.940 --> 00:06:01.500
had complied with the European Union's GDPR regulations.

124
00:06:01.500 --> 00:06:03.120
One of Suprema's clients

125
00:06:03.120 --> 00:06:05.580
was the United Kingdom's Metropolitan Police,

126
00:06:05.580 --> 00:06:07.740
and at the time of the breach,

127
00:06:07.740 --> 00:06:11.130
the United Kingdom was still part of the European Union.

128
00:06:11.130 --> 00:06:13.440
Furthermore, since the data was exposed

129
00:06:13.440 --> 00:06:15.270
across multiple countries,

130
00:06:15.270 --> 00:06:17.580
data sovereignty and the privacy laws

131
00:06:17.580 --> 00:06:20.640
of a number of countries were in effect.

132
00:06:20.640 --> 00:06:24.270
So remember, privacy risk is the potential

133
00:06:24.270 --> 00:06:27.180
for unauthorized access, misuse,

134
00:06:27.180 --> 00:06:29.670
or disclosure of personal data.

135
00:06:29.670 --> 00:06:32.250
Privacy risks are especially important

136
00:06:32.250 --> 00:06:36.180
to consider when dealing with technologies like biometrics,

137
00:06:36.180 --> 00:06:39.060
where unique personal data such as fingerprints

138
00:06:39.060 --> 00:06:41.760
and facial recognition are involved.

139
00:06:41.760 --> 00:06:45.180
Next, data subject rights provide individuals

140
00:06:45.180 --> 00:06:48.030
with control over their personal information,

141
00:06:48.030 --> 00:06:51.990
allowing them to access, correct, or delete this data.

142
00:06:51.990 --> 00:06:54.870
These types of rights are provided by regulations

143
00:06:54.870 --> 00:06:57.720
like GDPR and CCPA.

144
00:06:57.720 --> 00:07:00.750
Finally, data sovereignty is the principle

145
00:07:00.750 --> 00:07:02.880
that data is governed by the laws

146
00:07:02.880 --> 00:07:06.270
of the country in which it is collected or stored.

147
00:07:06.270 --> 00:07:09.450
So data sovereignty adds complexity

148
00:07:09.450 --> 00:07:13.473
for multinational companies who collect and store data.