Law and Policy in Responsible Technology

Law, Standards, and Policy in Responsible Technology

• This unit covers legal areas relevant to computer science.
• Topics of concern include privacy and data protection, cybercrime, intellectual property, copyright, equality legislation, and how computer science impacts law.
• The discussion also touches on supporting judicial decisions.
• It is important to understand the difference between ethics and the law, especially in regard to innovation outpacing regulation.
• Law protects essential parts of constitutional democracy.
• Knowledge of law is crucial for developers of computational systems, as their creations may shape legal interpretation.
• Computer systems can potentially weaken human rights or make legal remedies less effective.
• Laws are binding for all members of society, but dissent or disobedience may be justified for ethical considerations.
• Laws can sometimes be unethical.
• Policy is a general statement of intent for a government or prospective government.
• Policy is often enacted through legislation.
• Legislation (primary legislation) is law created by a central authority that members of the society are expected to follow.
• Regulation (secondary legislation) implements primary legislation, often created by domain experts or a relevant regulator within a specific sector.
• Standards are created by organizations or communities, either formally through expert consensus (de jure) or through common practice (de facto).
• Standards can inform legislation.
• Examples of legislative areas of concern include privacy and data protection, cybercrime, intellectual property, and equality legislation, and how technology impacts them.

Privacy and Data Protection

• The General Data Protection Regulation (GDPR) is relevant legislation, with further details available for independent study.
• Data collection should be limited to necessary data.
• Data should only be used for its intended purpose.
• Data should be accurate and current
• Users should be able to access and control their data.
• Data security is a vital aspect.

Cybercrime

• Cyber-dependent crimes rely on ICT for their actions.
• Cyber-enabled crimes utilize ICT to increase the scale or impact of traditional crimes.
• Ex: fraud, intellectual property crime (counterfeiting and copyright infringement).

Intellectual Property and Copyright

• Automated tools search for infringements related to stock photos and code snippets.
• Using code libraries and packages requires careful consideration of licenses (and track whether they are compliant).
• Important to consider the appropriate licenses of open-source software, particularly when it comes to commercial use.
• Use of software often requires consideration of attribution.

Equality Legislation

• Equality legislation aims to ensure people are treated fairly.
• It addresses prohibited behaviours, such as discrimination.
• The Equality Act of 2010 is key legislation, consolidating various prior acts (sex discrimination, race relations, and disability discrimination acts) to provide protection from various types of discrimination.
• Protected characteristics covered (e.g. disability, age, gender, religion or belief, sexual orientation, and marriage/civil partnership status) are addressed extensively in legislation to prevent discrimination and harassment.

Case Study - Amazon Recruitment Tool

• Amazon's recruitment tool used historical data and AI which favored male candidates.
• This highlights how algorithms can reflect biases present in data, leading to discrimination.

Application of IT to the Legal Process - COMPAS

• COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) is a tool used in the legal process to assess recidivism risk.
• Critically, the tool showed bias in its predictions, with disproportionate higher risk scores for Black individuals.
• Flaws in the algorithms revealed the inherent risk of algorithmic bias in decision-making contexts.

Algorithmic Bias - COMPAS

• Opacity of algorithms is a fundamental challenge.
• Laws usually undergo review and examination.
• Proprietary algorithms limit review opportunities; the lack of transparency presents challenges.
• Algorithmic biases violate equality legislation.
• Technological influence on law is a broader issue requiring further examination.

The Influence of Tech in Law

• Laws are created with checks and balances.
• Ethics can be influenced and established by tech firms within their organizations and use cases.
• Tech companies have the ability to enforce ethical choices that users should be able to understand.
• There can be issues with oversight and avenues for disobeying or dissenting in the context of tech.
• Users lack explicit control over ethical implications and may be unknowingly supporting certain decisions that they would not support explicitly.
• Firms like Microsoft and Amazon have withdrawn from market due to ethical concerns around the police use of facial recognition (which raises several ethical concerns).
• Who defines what is ethically permissible if tech firms dominate the definition of ethical behaviour?