Ethical and Legal Issue in AI and Data Science

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 13

COURSEWORK # 2

Student Name
Contents
Ethical and legal issue in AI and data science.................................................................................1

How else should We Address Income inequality?.......................................................................2

Can technologies affect how we react and respond?...................................................................2

Can AI bias be destroyed?...........................................................................................................3

Challenges in AI..............................................................................................................................3

Case study in AI and its solution.....................................................................................................4

Ethical dilemma...........................................................................................................................4

Ethical Case study Health Care.......................................................................................................5

Application of AI for health care.................................................................................................5

Ethical challenges........................................................................................................................5

International Law.........................................................................................................................5

Why is accountability required?..................................................................................................6

Social issue in AI.............................................................................................................................6

Policy...........................................................................................................................................6

Conciseness and legibility............................................................................................................7

Case study in AI and its solution.....................................................................................................8

Plot research in AI........................................................................................................................8

Legal issue in AI..............................................................................................................................9

Challenges in AI.........................................................................................................................10

Conclusion.....................................................................................................................................10

Reference.......................................................................................................................................11
Ethical and legal issue in AI and data science
Every time data is used to anticipate outcomes and help decision-making processes, people may
be affected in a variety of ways (Barocas & Selbst, 2016). The topic of ethical issues and the
“appropriate” use of data has only recently begun to attract the recognition it deserves, despite
the fact that that the rising field of data science has opened up many new potentials for problem-
solving and creating unique insights data - based (Saltz & Dewar, 2019). Since there doesn't
appear for a general agreement on what is judged legal and what is immoral a more detailed
examination is needed for said data science community. In terms of ethics, one must consider not
only the rights of individuals but also the rights of data collected from them.

The study of morality or the moral assessment of decisions are two definitions of ethics. Ethics
in data science serve as an example of right, acceptable, fair, but also excusable manner for
carrying out studies using this kind of data. Three main ethical theories serve as the foundation
for ethics: this same Immanuel approach (Louden, 1986), the utilitarian viewpoint (Shaw, 1999),
and the virtue model (Slote, 1992). "Pending the results, actions are to be viewed as morally right
or wrong, just or unjust, in themself".

Intelligent machine systems are improving our lives on a daily basis. This world grows more
economic greater capable computer systems grow.

Some of the biggest names in technology today think artificial intelligence (AI) has to be used
more frequently. To make this happen, there is a plethora of ethical and risk assessment
challenges that need to be kept in mind. These are covered below.

How else should We Address Income inequality?


Most individuals pay the bulk of their waking hours merely will have enough money to sustain
their and their families. The success of AI will also provide individuals the chance to spend more
time managing for their households and discovering new ways to add to humankind given the
amount of time it saves.

Can technologies affect how we react and respond?


Artificial intelligence (AI) bots are getting better at emulating interpersonal interactions and
discourse. In 2015, Eugene Grossman became the second machine to ace the Exam. In order to
accomplish this challenge, real raters will need to use text string to communicate with an
unknown entity and determine if it is a computer or a person. While this may be highly helpful in
influencing society to behave well, it can also be dangerous if used improperly.

Can AI bias be destroyed?


Don't neglect that individuals, who are typically exceedingly skewed and judging, are the ones
who design AI systems. Yes, AI has the capacity to spur progress if properly applied, but it may
also reinforce prejudice. The ability and speed that Summary—level considerably outperform
those of beings, yet because of anthropogenic influence, it can't only be relied upon to be
impartial and fair

Challenges in AI
A previous analysis of the ethical issues posed by AI has highlighted six categories of worries
that may be linked to the structure and operation of AI systems and decision-making algorithms.
The replica but modified map in Figure considers: “Judgment algorithms (1) transform data into
evidence for a certain conclusion (forevermore finding), and that this conclusion is then used to
(2) trigger and inspire an action that (on its own, or when paired with other acts, may not be
ethically neutral. The complicated yet (50 % replacement strategies employing this work make it
difficult to attribute blame for the results of analysis tool activities.

Predicated on how systems analyze data should provide findings and encourage actions, three
factual as well as two ideal varieties of ethics may be determined from these operational features.
The five identified sorts of concerns can lead to failures integrating several technological,
organizational, including human agents. This combination of technological and human actors
raises difficult issues on how to allocate blame and culpability for the consequences of AI
behaviors. Traceability integrates these issues as a final, all-encompassing sort of issue.

1. Business intelligence and mechanical outcome rely on patterns and logical knowledge.
Correlations built on a “sufficient” amount of data are frequently considered reliable
enough to guide action without demonstrating causality explicitly. Even if substantial
associations or causal insights are discovered, these findings may only apply to groups
while activities that have a major bearing on an individual's life are focused towards that
person.
2. Information asymmetry in machines learning algorithms is a result of the data's high
degree of dimension, advanced software, and flexible logic. The logic that translates
inputs into conclusions could not know to bystanders or other parties who are harmed by
it, or it could fundamentally illogical even illogical, which is the “black box” challenge
with AI. It is typically preferred that an algorithm be transparent and accessible, since
weakly foreseeable or easy to interpret methods are tough to regulate and improve.
3. Beliefs in Ai are a potent candidate of bias against by persons and groups. Discriminatory
analytics can contribute to stigmatization and self-fulfilling prophesies. In this situation,
it can be extremely complex to include fairness and of pas factors into Ai technologies. It
could be able to instruct algorithms to ignore delicate characteristics that support
discrimination.
4. It is particularly difficult under this context with intelligent systems, such as
recommendation systems, to adapt information. Algorithms' value-laden judgements may
potentially be a danger to autonomy. By removing content that is regarded irrelevant or in
conflict with the user's ideas or wishes, tailoring decreases the diversity of facts users
encounter.
5. Informational privacy refers to such a person's capacity to safeguard by her own private
information and or the time involved for outside parties to gain it. This might include
insurers, remote care providers, consumer technology firms, and others in the healthcare
industry. Data subjects find it difficult to create fundamental privacy rules that apply to
all forms of data.
6. Developers and implementation plan have historically had "complete control of the
behavior of the technology." This classic view of responsibility in software design makes
the assumption that such producer can think through the potential implications and flaws
of the equipment and make design to get the best desirable results.

Case study in AI and its solution


Ethical dilemma
The majority of said ethical and moral questions we find in life have somewhat simple solutions,
irrespective of the fact that we encounter many of them. An ethical difficulty, also known as
more than just a moral question or ethical ambiguity, arises when a individual has to decide
between two things, each of which is not wholly ethically acceptable.

Ethical Case study Health Care


If appropriately utilized healthcare professionals might concentrate on the underlying causes of
disease and monitor the efficacy of therapies as well as preventative measures. A key component
of health policy, per the computer and information specialists and hospital entrepreneurs, will be
artificial intelligence (AI), particularly machine learning. The ability of a programming language
to put on functions that seem to be in line with human intellect is referred to as AI in the
computing industry. Overall healthcare settings, has been a simultaneous increase of data.
Simply described, traditional computational algorithms are computer systems that constantly
perform identical work following a set of rules, like an electronic calculator that says "if this is
the input, then this is the output".

Application of AI for health care


An important piece of AI-based healthcare research is the use of data produced for health.
Therapeutic practices may be found with the use of AI that has been functioning optimally and
taught with sufficient data. AI may also help in the development of future therapeutic
practitioners of health services by studying specific clinical patterns revealed through electronic
health data.

Ethical challenges
A question like whether AI "fits within current legal groups or whether a new category with its
particular features and implications should be formed" is one that is constantly being discussed
(Resolution of the European Parliament, 16 February 2017). Data privacy, analytical unfairness
but also limitations, full access to use of data, and these four fundamental ethical challenges must
all be addressed.

International Law
Based on a request from the House on Public Affairs of the European Parliament, this same
policy unit for "Individuals' Privileges and Basic Concerns" commissioned, oversaw, and
published the study that served as the foundation for the resolution.
Why is accountability required?
Intelligent systems are exposed of sudden, severe failures when the environment or context
changes. Even if AI discrimination is mitigated, all AI systems will have their limitations. The
system's limits must also be understood by the strategic decision. Concerns regarding cyber
security flaws are raised when AI is deployed without human input. A threat vector based on
"data diet" risks is developed when AI is used in cyber security or surveillance, according to a
RAND views research.

Social issue in AI
Policy
Laws and other forms of legislation govern human society. The development of AI and machine
learning has often avoided legal issues while remaining in the academic domain, but somehow
the future

While these technologies begin to dominate society world at large, their effects on individuals
are likely to run into certain impediments. The use of AI as the foundation for autonomous
driving cars is one hotly debated example. Legal obligations are also very strictly outlined
whenever a supervisor is in control of almost any judgement while operating a vehicle. However,
the present legal framework is becoming increasingly strained as a result of the rapidly
industrialization of de facto independent cars that is moving us closer to the goal of completely
autonomous navigation. Any use of AI and ML in real-world medical management is certain to
spark debate about its ethical implications. The Data Protection Act Regulation's recent (May
2018) activation by the European Union serves as a prime illustration (GDPR). The "data
master" is legally obliged to explain to the public any actions taken by "automatically or
synthetic intelligent algorithmic algorithms" in in accordance with Article 13 of the directive.

Conciseness and legibility


The ability to articulate oneself has not certainly evolved in human brains. It might be believed
that social behaviors can only occur in creatures with brains that are capable of self-explanation
through language, and that this has only occurred in mammals with behavioral. Natural language
serves that interpretive or communicative purpose in the rest of humanity.

Self-explanatory skills was never a major factor to account for once it was first conceptualized as
an attempt to imitate parts of biological intelligence. The study topic of interpretability and
modifiability of AI and ML platforms has just lately moved to the fore. DL models run the
danger of being viewed as enhanced black boxes. This has significant consequences for biology
since it can exacerbate the issue of "the need to open its deep learning algorithm black box" if an
Ounces MDSS makes a mop legible. Medical professionals frequently view computer-based
medical diagnosis and prognosis tools (MDSS) as an additional burden in their daily work.

When the MDSS clashes with standards of medical practice, a problem may arise. It should be
made an attempt to incorporate medical experts.

Challenges in AI

1. One cause why developers avoid the subject is the amount of power that these algorithms
need. Deep learning frameworks may be used in a wide range of fields, namely asteroid
tracking, the delivering of healthcare, the detection of celestial bodies, and many more.
2. Ai technology may be used in a variety of markets as a superior replacement for current
technologies. Only a small percentage of individuals, excluding computer enthusiasts,
college students, and researchers, are aware of the possibilities of AI. The understanding
of a.i. has been the true issue.
3. One of the most serious AI issues, it has kept academics on their toes for businesses and
start-ups. Although these businesses may brag of accuracy greater rates 90%, employees
could do better in each of these situations. Let our model, for examples, determine if the
image is of a dog or a cat. A typical being can foresee the correct outcome almost every,
with an astonishing accuracy of above 99%.

Case study in AI and its solution


Plot research in AI
Ai systems (AI) is being more commonly regarded in society as a result of a series of
technological developments, symbolizing the beginning of AI's emerge as a viable, useful, and
revolutionary technology. In the American series Katya!, IBM's Williams took first place, and
Google's Human players defeated the Go world champion. Such encouraging achievements give
hope for an exciting future made possible by the fusion of artificial intelligence technology and
human ingenuity. The study but instead promotion of AI technology has received major funding
from nations all over world. IBM is also using AI to try to stop the Ziska virus from spreading.

Especially ASEAN has rising mortality rates, as opposed to such OECD regions' dropping
suicide rates. Between OECD nations, Korea current has the highest suicide rate. The study aims
to: (1) explain suicidal urge by examining societal traits of people who have experienced suicidal
impulse; (2) forecast the possibility of attempted suicide; and (3) assess spatially quality of life.
The study of clinical risk special obligations use of data from the Korean Participation and the
Survey of Youth Health History. Since there are not enough people trying to commit suicide,
there are too many factors that might influence it. To pinpoint subgroups at risk of suicide
attempts, the concepts of support, certainty, and volume are proposed.

Legal issue in AI
A legal issue is a scenario which could require the assistance of something like a lawyer to
resolve because it has potential legal ramifications. A legal issue is a situation which might
require the help of a lawyer to remedy because it has probable legal ramifications. Immigration
and asylum, consumer rights, especially housing challenges are all significant constitutional
implications. Users can deal various situations more adequately if you are aware of the rules as
well as your rights. There are many various ways that legal concerns might arise, including from
anticipated occurrences like purchasing a property or creating a will.
Challenges in AI
In 2018, the number of new machine learning-based tools, initiatives, and apps increased
dramatically. The legal, medical, transportation, financial, and many more fields will be
impacted by this. A bystander was killed when a car accident committed by an Uber self-driving
vehicle, and raised the issue of culpability. The Cambridge Analytical incident also served as a
turning point for AI. In 2020, autonomously buses will begin operating on Swedish highways.

Conclusion
We are at the beginning of a shift in this rapidly growing world of automated vehicles, medical
and health care robots, sophisticated bots, and other innovation. We must carefully approach the
new technology and take into account all its legal and ethical ramifications. The EU Parliament
has previously created an ethics code to address the ethical obligation of engineers and
researchers. We have looked at some of the most blatant societal and legal issues that artificial
intelligence raises. We specifically looked at social difficulties as usually difficult circumstances.
The subject question of whether we should anticipate pro-social action from artificial intelligence
would have no simple or broadly supported solution. It depends on the facts and the legal
implications of either conclusion, as is so frequently the case.
Reference

[1]. Stahl, B. C. (2021). Ethical issues of ai. In Artificial Intelligence for a Better Future (pp. 35-53).
Springer, Cham.
[2]. Ouchchy, L., Coin, A., & Dubljević, V. (2020). AI in the headlines: the portrayal of the ethical
issues of artificial intelligence in the media. AI & SOCIETY, 35(4), 927-936.
[3]. Siau, K., & Wang, W. (2020). Artificial intelligence (AI) ethics: ethics of AI and ethical
AI. Journal of Database Management (JDM), 31(2), 74-87.
[4]. Ghotbi, N., Ho, M. T., & Mantello, P. (2022). Attitude of college students towards ethical issues
of artificial intelligence in an international university in Japan. AI & SOCIETY, 37(1), 283-290.
[5]. Stahl, B. C. (2021). Addressing Ethical Issues in AI. Artificial Intelligence for a Better Future,
55-79.
[6]. Bostrom, N. (2003). Ethical issues in advanced artificial intelligence. Science fiction and
philosophy: from time travel to superintelligence, 277, 284.
[7]. Cath, C. (2018). Governing artificial intelligence: ethical, legal and technical opportunities and
challenges. Philosophical Transactions of the Royal Society A: Mathematical, Physical and
Engineering Sciences, 376(2133), 20180080.
[8]. Karliuk, M. (2018). Ethical and Legal Issues in Artificial Intelligence. International and Social
Impacts of Artificial Intelligence Technologies, Working Paper, (44).
[9]. Stahl, B. C. (2021). Ethical issues of ai. In Artificial Intelligence for a Better Future (pp. 35-53).
Springer, Cham.
[10]. Siau, K., & Wang, W. (2020). Artificial intelligence (AI) ethics: ethics of AI and ethical
AI. Journal of Database Management (JDM), 31(2), 74-87.
[11]. Stahl, B. C., Antoniou, J., Ryan, M., Macnish, K., & Jiya, T. (2022). Organisational
responses to the ethical issues of artificial intelligence. AI & SOCIETY, 37(1), 23-37.
[12]. McLennan, S., Fiske, A., Celi, L. A., Müller, R., Harder, J., Ritt, K., ... & Buyx, A.
(2020). An embedded ethics approach for AI development. Nature Machine Intelligence, 2(9),
488-490.
[13]. Rodrigues, R. (2020). Legal and human rights issues of AI: Gaps, challenges and
vulnerabilities. Journal of Responsible Technology, 4, 100005.
[14]. Price, I. I., & Nicholson, W. (2017). Artificial intelligence in health care: applications and
legal issues.

You might also like