1
00:00:00,050 --> 00:00:00,740
Case study.

2
00:00:00,740 --> 00:00:03,050
Navigating global AI regulations.

3
00:00:03,050 --> 00:00:06,470
Tech Nova's journey in ethical and compliant AI development.

4
00:00:06,500 --> 00:00:12,410
AI regulations are essential for ensuring ethical practices, compliance and fostering public trust

5
00:00:12,410 --> 00:00:15,740
in technologies that are rapidly shaping modern industries.

6
00:00:16,100 --> 00:00:22,190
At Tech Nova, a leading AI driven healthcare startup, the implications of AI regulations became evident

7
00:00:22,190 --> 00:00:28,010
when the company decided to implement a new AI diagnostic tool intended to revolutionize patient care

8
00:00:28,010 --> 00:00:31,280
by providing real time analysis of medical images.

9
00:00:31,700 --> 00:00:38,690
The team, led by CEO Doctor Eliza Carter, included AI engineers, legal advisors and ethical officers

10
00:00:38,690 --> 00:00:40,970
who were well versed in AI governance.

11
00:00:41,810 --> 00:00:47,450
The primary challenge faced by Tech Nova was ensuring that their AI diagnostic tool complied with the

12
00:00:47,450 --> 00:00:50,720
General Data Protection Regulation in the European Union.

13
00:00:51,410 --> 00:00:58,700
Given that the tool processed sensitive patient data ensuring GDPR compliance was non-negotiable, the

14
00:00:58,700 --> 00:01:04,940
team was particularly concerned about article 22 of the GDPR, which restricts decisions based solely

15
00:01:04,970 --> 00:01:10,580
on automated processing without human intervention, how could Tech Nova integrate meaningful human

16
00:01:10,580 --> 00:01:14,870
oversight while retaining the efficiency and accuracy of their AI tool?

17
00:01:16,730 --> 00:01:21,440
Doctor Carter initiated a series of meetings with her team to brainstorm solutions.

18
00:01:21,950 --> 00:01:27,920
They decided to incorporate a system where AI generated diagnoses would be reviewed by medical professionals

19
00:01:27,920 --> 00:01:30,140
before being communicated to patients.

20
00:01:30,590 --> 00:01:37,190
This not only ensured compliance with GDPR, but also added a layer of human judgment to the AI's conclusions.

21
00:01:37,550 --> 00:01:43,400
However, this raised another question would the added layer of human oversight compromise the tool's

22
00:01:43,400 --> 00:01:44,360
efficiency?

23
00:01:45,830 --> 00:01:51,620
To address this, the team conducted a pilot study comparing diagnosis times and accuracy rates between

24
00:01:51,620 --> 00:01:55,040
the AI only and AI plus human oversight models.

25
00:01:55,280 --> 00:01:56,900
The results were promising.

26
00:01:56,900 --> 00:02:02,510
While the oversight added a minimal time delay, it significantly increased the accuracy and reliability

27
00:02:02,510 --> 00:02:03,710
of diagnoses.

28
00:02:04,160 --> 00:02:09,890
This approach balanced innovation with regulatory compliance, enhancing patient trust and safety.

29
00:02:11,240 --> 00:02:17,330
In parallel, Terranova faced the challenge of navigating the proposed EU Artificial Intelligence Act,

30
00:02:17,330 --> 00:02:20,540
which categorized AI systems by risk levels.

31
00:02:20,960 --> 00:02:26,690
Their diagnostic tool, being a high risk application, required stringent regulatory measures, including

32
00:02:26,720 --> 00:02:28,970
transparency and risk assessments.

33
00:02:29,540 --> 00:02:34,760
The team developed a comprehensive risk management framework drawing upon guidelines from the National

34
00:02:34,760 --> 00:02:36,920
Institute of Standards and Technology.

35
00:02:37,670 --> 00:02:43,250
The NIST AI Risk Management Framework provided a structured approach to identify, assess, and manage

36
00:02:43,250 --> 00:02:45,320
risks associated with their tool.

37
00:02:45,350 --> 00:02:50,450
But how could Technova ensure continuous compliance given the evolving AI landscape?

38
00:02:52,160 --> 00:02:57,620
This prompted the implementation of an ongoing monitoring system, where the AI tools performance was

39
00:02:57,620 --> 00:03:01,400
continuously evaluated and adjustments were made as needed.

40
00:03:01,690 --> 00:03:06,970
Regular audits were conducted to assess compliance with both GDPR and the AI act.

41
00:03:07,810 --> 00:03:13,570
This proactive approach not only ensured compliance, but also provided valuable data for further improving

42
00:03:13,570 --> 00:03:16,600
the AI system across the Atlantic.

43
00:03:16,630 --> 00:03:22,960
Terranova had to adapt its strategy to align with the United States sector specific AI regulations.

44
00:03:23,530 --> 00:03:29,560
The Federal Trade Commission guidelines on AI and consumer protection emphasized fairness and transparency.

45
00:03:29,680 --> 00:03:35,020
Given the health care sector's vulnerability to biased outcomes, the team undertook a comprehensive

46
00:03:35,020 --> 00:03:40,480
review of their AI models to identify and mitigate any potential biases.

47
00:03:40,480 --> 00:03:46,540
How could they ensure that their AI tool provided equitable care to all patients, regardless of demographic

48
00:03:46,540 --> 00:03:47,350
factors?

49
00:03:48,580 --> 00:03:54,520
The team conducted an extensive bias audit, analyzing the model's performance across different demographic

50
00:03:54,520 --> 00:03:55,330
groups.

51
00:03:55,810 --> 00:04:01,720
They implemented adjustments to address identified biases, ensuring the AI provided consistent and

52
00:04:01,720 --> 00:04:03,180
fair diagnoses.

53
00:04:03,210 --> 00:04:08,910
Additionally, they developed a transparent explanation mechanism allowing patients and health care

54
00:04:08,910 --> 00:04:12,540
providers to understand the AI's decision making process.

55
00:04:12,570 --> 00:04:18,570
This enhanced transparency fostered greater trust and acceptance of the AI tool among users.

56
00:04:19,530 --> 00:04:25,350
Meanwhile, Terranova's expansion into the Chinese market brought another layer of regulatory complexity.

57
00:04:25,950 --> 00:04:31,770
The Chinese government's AI regulations, as outlined in the New Generation Artificial Intelligence

58
00:04:31,770 --> 00:04:36,990
Development Plan, emphasized ethical principles and stringent data protection standards.

59
00:04:37,620 --> 00:04:41,340
How could Tekmar navigate these regulations while promoting innovation?

60
00:04:43,470 --> 00:04:48,870
Terranova collaborated with local experts to ensure compliance with Chinese regulations, including

61
00:04:48,900 --> 00:04:53,220
obtaining necessary licenses and adhering to data protection standards.

62
00:04:53,880 --> 00:04:58,830
The team modified their data handling practices to align with local requirements, ensuring patient

63
00:04:58,830 --> 00:05:02,160
data was stored and processed securely within China.

64
00:05:02,550 --> 00:05:08,790
This localization strategy not only ensured compliance, but also facilitated smoother market entry

65
00:05:08,790 --> 00:05:10,980
and acceptance of their AI tool.

66
00:05:12,840 --> 00:05:18,780
Beyond national regulations, international organizations like the organization for Economic Cooperation

67
00:05:18,780 --> 00:05:23,610
and Development played a crucial role in shaping Tech Nova's AI governance strategy.

68
00:05:24,240 --> 00:05:30,900
The OECD principles on AI provided a global framework for responsible AI development, emphasizing human

69
00:05:30,900 --> 00:05:33,030
centered values and accountability.

70
00:05:33,750 --> 00:05:38,220
How could Tech Nova integrate these principles into their AI governance framework?

71
00:05:39,750 --> 00:05:45,360
The team developed a comprehensive AI governance policy incorporating the OECD principles.

72
00:05:45,450 --> 00:05:51,660
This policy emphasized inclusive growth, sustainable development, and human centered values in all

73
00:05:51,660 --> 00:05:52,830
AI initiatives.

74
00:05:53,850 --> 00:05:58,860
Regular training sessions were conducted to ensure that all employees understood and adhered to these

75
00:05:58,860 --> 00:06:04,140
principles, fostering a culture of ethical AI development within the organization.

76
00:06:05,040 --> 00:06:10,830
Despite these efforts, the global nature of AI development posed challenges in harmonizing regulations

77
00:06:10,830 --> 00:06:12,450
across jurisdictions.

78
00:06:12,810 --> 00:06:18,810
Discrepancies in regulatory approaches created barriers to innovation and complicated compliance efforts

79
00:06:18,810 --> 00:06:19,800
for technova.

80
00:06:20,280 --> 00:06:26,430
How could the company navigate these challenges while promoting a cohesive global framework for AI governance?

81
00:06:28,140 --> 00:06:33,720
Technova joined the Global Partnership on AI, an international initiative promoting responsible AI

82
00:06:33,750 --> 00:06:36,120
development through GPI.

83
00:06:36,150 --> 00:06:42,690
Technova engaged in dialogue with governments, industry and academia, sharing best practices and collaborating

84
00:06:42,690 --> 00:06:44,190
on AI research.

85
00:06:44,820 --> 00:06:50,460
This facilitated the development of common standards for AI governance, helping bridge the gap between

86
00:06:50,460 --> 00:06:52,440
different regulatory approaches.

87
00:06:53,250 --> 00:06:59,070
In conclusion, Technova journey highlights the importance of understanding and navigating AI specific

88
00:06:59,070 --> 00:07:01,800
regulations across different jurisdictions.

89
00:07:02,280 --> 00:07:08,000
The company's proactive approach to integrating human oversight, conducting bias audits, and adopting

90
00:07:08,000 --> 00:07:15,260
international principles, ensured compliance, and fostered ethical AI practices by engaging in international

91
00:07:15,260 --> 00:07:16,310
collaborations.

92
00:07:16,310 --> 00:07:22,100
Terranova contributed to the development of a cohesive global framework for AI governance.

93
00:07:22,100 --> 00:07:27,530
These efforts underscore the need for continuous adaptation and vigilance in the evolving landscape

94
00:07:27,530 --> 00:07:32,210
of AI regulations for AI governance professionals.

95
00:07:32,210 --> 00:07:38,000
Terranova's case study offers valuable insights into the practical application of regulatory frameworks

96
00:07:38,000 --> 00:07:39,710
and ethical principles.

97
00:07:40,100 --> 00:07:46,580
It emphasizes the importance of proactive risk management, ongoing compliance, and international collaboration

98
00:07:46,580 --> 00:07:51,230
in navigating the complex and multifaceted landscape of AI regulations.

99
00:07:51,800 --> 00:07:57,980
By learning from Nova's experiences, Agep professionals can enhance their understanding and application

100
00:07:57,980 --> 00:08:03,860
of AI governance, ensuring responsible and ethical AI development in their respective organizations.
