Artificial Intelligence (AI) & AI Literacy

Generative AI tools like ChatGPT, DALL-E, and more are trained and managed via large data centers. As of April 2025, the U.S. houses over 3,600 data centers, with about the most being in Northern Virgina, Texas, and California. A high amount of electricity powers these robust data centers, and, in turn, requires copious amounts of freshwater to cool them down. Most data centers use fresh, drinkable water for the Hereculian task of cooling such powerful systems, utilizing tubes surrounding IT equipment. According to Bloomberg News, two-thirds of data centers built since 2022 are located in areas of crippled by high levels of water stress, thus further exhausting already-depleted areas. 

Corporations like Microsoft and Google note their contribution to carbon emissions is increasing due to construction of new data centers to support AI. The demand for electricity to run these AI data centers contributes to rising carbon dioxide emissions and more demand to electric grids. 

Exhaust from data centers can emit harmful toxins into the immediate area. For example, in the American South, Memphis, a majority-Black city, is currently reckoning with harmful pollutants like smog, nitrogen oxides, and formaldehyde contaminating the air. This is thanks in part to Elon Musk’s xAI, a powerful data warehouse, whose unpermitted turbines run constantly, continuing to plague the community. 

It’s worth highlighting that organizations and advocacy groups have called for steps to be taken to make AI greener. Proposals for moves in the right direction include the alternative of running cloud data centers on carbon-free energy by 2030, utilizing hardware that releases less heat, and more. Steps in the right direction for the sake of our environment must be taken now, and users should hold these corporations accountable for the commitments they are making.

The video below is a clip from PBS NewsHour that talks about the growing effect AI data centers have on the environment, shedding light on the fact that there are currently (at the time this guide was created in 2025) very little regulations in place to mitigate the use of fossil fuels, energy, and water consumption. 

It is crucial to be aware that there are biases present in AI software, particularly because AI has been created by humans who have internal biases. This is especially apparent in facial recognition software and image generating AI tools that have either failed to recognize faces of color, erroneously mistaken one person for another, or have generated images that reveal a problematic pattern when inputting specific queries (typing characteristics for a lawyer, for instance, revealed mostly white, male figures). 

ChatGPT and, most recently, xAI's chatbot, Grock, have also been known to iterate hate speech, racial biases, as well as gender stereotypes, from writing resumes, producing recommendation letters, constructing college major recommendations, and even in drawing up medical reports. Microsoft’s Tay, an AI chatbot, had been released through Twitter (now X) back in 2016 and was removed within 24 hours because of multiple racist, sexist, anti-semitic, and false information that had been posted. It bears stating again that, because LLMs are trained and programmed by humans with implicit biases, tools like ChatGPT and Tay will draw on the information that is available on the internet, whether or not the information is factually accurate or contains problematic language. Before completely relying on AI software and other algorithmic tools, it is imperative to check and pay close attention to accuracy, in addition to questionable wording or phrasing that might reveal harmful rhetoric against marginalized communities. 
The videos below feature Dr. Joy Buolamwini, MIT researcher and founder of the Algorithmic Justice League, and Dr. Safiya Noble, the David O. Sears Presidential Endowed Chair of Social Sciences, Director of the Center on Resilience & Digital Justice, and UCLA professor, whose research both emphasize the need for ethical integration in AI technology, stressing that – as it currently stands – GenAI must incorporate more equitable programming practices and policies to reduce and minimize harm towards women and people of color.

While advancements in AI technology have provided benefits such as more streamlined workflow processes, assistance with language learning, and visual/artistic enhancement capabilities (among others), privacy concerns–as well as data breaches–should be taken into consideration. Whatever input is added into a GenAI tool (whether it's through ChatGPT or another system), is often saved and stored into the company that runs the system in order to analyze the information and then train their models to improve them.

GenAI has also been exploited for scams and fraud through voice phishing and identity theft by creating believable text, images, and even voice mimicry for the purpose of impersonating celebrities or even members of the public to extort funds from organizations and other individuals. There have been a number of instances reported where AI-generated images have also been used in photos, videos, and other content without the subject’s consent or knowledge. These cases are not limited to well-known public figures and have affected communities globally.

AI-driven phishing scams are another example of data breaches caused by GenAI tools. These attempts often target online accounts from platforms for wide-ranging services, including Gmail, healthcare/insurance, and even dating apps. As previously mentioned, AI is used in facial recognition technology for surveillance purposes, but has also been known to wrongfully identify individuals as criminals. Corporations and large companies have taken advantage of this tool to scrutinize their employees, which sometimes led to faulty data and mistaken identities. This often disproportionately affects Black and Brown folks since the software would incorrectly identify them, leading to unlawful arrests.   

The consequences of incorporating AI into healthcare – particularly the deleterious effects this has on the Black community who are most at risk and in need of medical attention – are mentioned in the video clip below by Sayash Kapoor, Princeton University’s Computer Science Ph.D. candidate and co-author of the book, AI Snake Oil.

Programs like ChatGPT and other LLMs are trained on information that is freely and readily available online. However, this includes original work that falls under copyright protection according to Title 17 of the U. S. Code § 201. ChatGPT’s creator, OpenAI, has been accused of violating copyright law and infringing on the privacy of a vast number of users because it employed data scraping techniques to train ChatGPT. 

Although there are a number of similar lawsuits currently underway which include tech giants like Google, Microsoft, Midjourney, and more (with cases ranging from written works, music, art, etc.), the most recent and notable development surrounding intellectual property and artificial intelligence is the ruling of two separate lawsuits involving AI companies, Anthropic and Meta. Writers and creators argued that these companies have trained their LLMs using the authors’ works without permission, which they believed counted as copyright infringement. However, both cases ruled in favor of Anthropic and Meta, claiming that training their large language models with the writers’ published materials was permissible under fair use. 

In addition to these allegations, both Anthropic and Meta had accessed and downloaded these works through pirated sites. The rulings of both cases raise important questions for courts to consider moving forward regarding AI’s role in creative content:

• What constitutes copyright for works that are available and uploaded online?
• Does the onus lie on platforms and pages that allow pirated content to be made readily available (and those who upload works protected by copyright), or on companies like Anthropic and Meta that access and use them for LLM training purposes?
• What are the implications for future human-AI collaboration?
• Will artificial intelligence eventually be recognized as an author, collaborator, or co-creator for works of art and/or scholarly publications?
• Or will the nature of AI and how it generates content prove to be a deterrent towards its acknowledgement and credibility as a creator?

The following videos provide brief overviews and perspectives on how the rulings on current court cases (at the time of this LibGuide’s creation in 2025) might possibly affect copyright regulations over different sectors of the arts and intellectual property: