Drafts with AI – Part 2 – Professor Karen Sichler

In this installment, I gave Perplexity Pro the following prompt:

Please write the body of a chapter on AI and privacy for a textbook titled Digital Rights and Privacy in the United States: A Primer. The textbook is for freshmen and sophomore university students primarily studying communication studies. You must use academic and industry sources to write it.
You must include the following points
1.) How does the technology and use of AI intersect with privacy?
2.) What are the current laws that connect with AI use and technology? United States? Canada? Mexico? EU? Africa? Asia?

I was surprised by the number of tasks this query created. This is how Perplexity Pro dissected the request.

Here is the output:

AI and Privacy: Challenges and Regulatory Landscapes

Artificial intelligence (AI) systems rely on vast datasets, often containing personal information, to train algorithms and make decisions. This intersection of AI and privacy raises critical concerns about data collection, surveillance, algorithmic bias, and transparency. Below, we examine how AI impacts privacy and analyze the evolving legal frameworks addressing these issues across key regions.

a. How AI Intersects with Privacy

Data Collection and Utilization
AI systems require massive datasets, frequently including sensitive personal information, to improve accuracy and functionality1 15. This raises risks of unauthorized access, reidentification of anonymized data, and misuse of information beyond its original purpose1 2. For example, generative AI models like large language models (LLMs) often scrape publicly available data, including personal details, without explicit consent2 11.
Surveillance Technologies
AI-powered tools such as facial recognition, biometric tracking, and predictive policing enable pervasive monitoring in public and digital spaces1 12. While these technologies can enhance security, they risk normalizing indiscriminate surveillance and eroding anonymity1 18.
Automated Decision-Making
AI-driven decisions in hiring, lending, and healthcare often lack transparency, making it difficult to contest outcomes or identify biases1 18. For instance, opaque algorithms may disproportionately harm marginalized groups by replicating historical inequities15 18.
Informed Consent Challenges
The complexity of AI systems complicates traditional consent models. Users may not fully understand how their data is used, undermining meaningful control over personal information1 2.

b. Current Laws Governing AI and Privacy

United States

Federal Level: No comprehensive AI law exists, but the 2023 White House Executive Order mandates safety assessments, anti-discrimination measures, and privacy protections for AI systems14 16. Sector-specific laws (e.g., HIPAA for healthcare) partially address AI use.
State Laws:
- California: The California Consumer Privacy Act (CCPA) grants residents rights to opt out of data sales and request explanations for automated decisions7 16. The 2026 AI Transparency Act will require disclosures about training data sources and content16.
- Colorado: High-risk AI systems must undergo impact assessments, provide consumer explanations, and mitigate algorithmic discrimination17.
- Illinois: Bans AI-driven hiring tools that analyze facial expressions without consent16.

North America

Canada
- Federal Bill C-27, stalled in 2025, proposed the Artificial Intelligence and Data Act (AIDA) to regulate high-risk AI systems5.
- Québec’s Law 25 mandates transparency for automated decisions, allowing individuals to request corrections or human review5 18.
Mexico
- No AI-specific laws exist, but the Ley Federal de Protección de Datos (LFPD) requires consent for personal data use. Proposed bills, like the 2024 National AI Law, aim to establish ethical guidelines and a regulatory body6.

European Union

EU AI Act: The world’s first comprehensive AI law bans unacceptable-risk systems (e.g., social scoring) and imposes strict transparency and safety requirements for high-risk applications (e.g., hiring tools)8 9. Phased enforcement began in 2025, with full compliance by 20278 19.
GDPR: Requires lawful data processing, consent, and explainability for AI decisions affecting individuals13 19.

Africa

Only 13 countries (e.g., South Africa, Kenya) have national AI strategies, but enforcement remains weak10 20. Existing data laws, like Nigeria’s NDPR, lack AI-specific provisions, leaving gaps in addressing algorithmic bias or automated decisions20.

Asia

Japan: Promotes an “AI-friendly” approach, easing consent requirements for data used in AI development11.
South Korea: Grants citizens rights to explanations and appeals for AI-driven decisions in employment and finance12.
Hong Kong: Requires ethical AI governance, including impact assessments and human oversight12.

Conclusion

AI’s rapid advancement challenges traditional privacy frameworks, necessitating adaptive laws that balance innovation with rights protection. While the EU leads with risk-based regulations, the U.S. relies on state-level laws, and regions like Africa and Asia face fragmentation. For communication students, understanding these dynamics is critical to navigating digital rights in an AI-driven world.

Citations:

Answer from Perplexity: pplx.ai/share

As you can imagine, the original draft provided by Perplexity AI left quite a lot to be desired. I then expanded the prompt.

Using the information provided in the most recent response, write the draft using a more traditional textbook style which includes paragraphs. Give specific examples of how the different laws will affect citizens in the various nations. Finally, compare the proposed laws to the current state of legislation in the United States of America.

Here is the second attempt:

AI and Privacy: Challenges and Regulatory Responses

The integration of artificial intelligence (AI) into daily life has transformed how personal data is collected, analyzed, and utilized, raising significant privacy concerns. AI systems depend on extensive datasets-often containing sensitive information-to train algorithms, automate decisions, and predict behaviors. This chapter explores the intersection of AI and privacy, examines global regulatory responses, and evaluates how these laws impact citizens, with a focus on comparisons to U.S. legislation.

AI’s Impact on Privacy

AI technologies challenge traditional privacy frameworks through pervasive data collection, surveillance, and opaque decision-making. For example, facial recognition systems deployed in public spaces can track individuals without consent, eroding anonymity and enabling mass monitoring. In healthcare, AI tools analyzing patient records risk exposing sensitive medical histories if data breaches occur. Generative AI models, such as ChatGPT, often scrape publicly available data-including social media posts or creative works-without explicit user permission, blurring ethical boundaries around data ownership. Automated hiring algorithms, trained on historical employment data, may perpetuate biases against marginalized groups, limiting opportunities for underrepresented applicants. These examples underscore the tension between AI innovation and individual privacy rights.

Global Regulatory Landscapes

United States
The U.S. lacks comprehensive federal AI legislation but relies on sector-specific laws and state initiatives. California’s Consumer Privacy Act (CCPA) enables residents to opt out of data sales and request explanations for AI-driven decisions, empowering individuals to contest discriminatory outcomes in credit or employment. Colorado’s AI Transparency Act (effective 2026) will require companies to disclose training data sources, allowing citizens to identify potential biases in tools like loan approval algorithms. At the federal level, President Biden’s 2023 Executive Order mandates safety testing for high-risk AI systems, aiming to prevent harms such as deepfake-generated misinformation. However, the fragmented approach leaves gaps; for instance, Illinois’ ban on emotion-recognition tools in hiring does not apply to other industries or states.

Canada
Canada’s proposed Artificial Intelligence and Data Act (AIDA), part of Bill C-27, would regulate “high-impact” AI systems, requiring transparency reports and risk mitigation. If enacted, citizens could demand explanations for automated decisions in sectors like banking, fostering accountability. Québec’s Law 25 already mandates that individuals receive clear information about AI-driven processes, such as resume screening tools, and request human intervention. These measures aim to balance innovation with trust but remain stalled in legislative debates.

Mexico
Mexico’s Federal Data Protection Law (LFPD) requires consent for personal data use but lacks AI-specific provisions. A proposed National AI Law (2024) seeks to establish ethical guidelines and a regulatory body, which could empower citizens to report unethical AI practices, such as biased policing algorithms. Current gaps, however, leave Mexicans vulnerable to unregulated surveillance technologies.

European Union
The EU’s AI Act (2025–2027 phased implementation) bans “unacceptable-risk” applications like social scoring and mandates transparency for high-risk systems. For example, employers using AI to evaluate job candidates must inform applicants and allow appeals, reducing opaque rejections. Coupled with the General Data Protection Regulation (GDPR), which requires consent and explainability, EU citizens enjoy robust recourse against algorithmic harms-a contrast to the U.S. patchwork.

Africa
Only 13 African nations, including South Africa and Kenya, have drafted AI strategies, but enforcement is inconsistent. Nigeria’s Data Protection Regulation (NDPR) lacks AI-specific rules, leaving citizens exposed to unchecked facial recognition systems used by private firms or governments. Without stringent safeguards, algorithmic bias in public services, such as welfare allocation, risks exacerbating inequalities.

Asia
South Korea’s AI Ethics Charter grants citizens the right to explanations for AI decisions in finance and employment, enabling challenges to unfair loan denials. Japan’s “AI-friendly” policies allow companies to use publicly sourced data without explicit consent, prioritizing innovation over individual control. In Hong Kong, mandatory ethical assessments for government AI projects aim to prevent misuse in policing or surveillance.

Comparison to U.S. Legislation

The EU’s risk-based regulatory model contrasts sharply with the U.S. approach. For instance, the AI Act’s outright ban on social scoring has no equivalent in U.S. law, where states like California focus narrowly on data transparency rather than systemic prohibitions. Similarly, Canada’s proposed AIDA mirrors the EU’s emphasis on high-risk system oversight, while U.S. federal efforts remain aspirational. Mexico’s nascent AI governance framework lags behind even the limited federal safeguards in the U.S., such as the White House’s non-binding AI Bill of Rights.

Conclusion

AI’s rapid evolution demands adaptive privacy protections that balance innovation with individual rights. The EU and Canada prioritize centralized, rights-based frameworks, whereas the U.S. relies on state-led experimentation. For communication students, understanding these disparities is critical to advocating for equitable policies in an increasingly automated world. As AI permeates media, advertising, and public discourse, transparent governance will shape whether technology empowers or endangers privacy.