Black communities, however, AI already shapes policing, media, and narratives. The urgent question is not if AI will influence society, but whether today’s systems will respect the lived experiences of those most affected. 

A recent national study by Jobs for the Future (JFF) reveals that Black learners and workers are among the most active users of AI tools. While 83% of Black respondents reported familiarity with AI, more than half also stated that they use AI weekly for learning, work, or creative projects (Swartsel, 2025). This frequent use reflects optimism about AI’s potential. Many saw promise in improved education and expanded career opportunities. Nevertheless, this optimism does not erase the fact that AI systems often mirror the inequities of the world that built them.
 

This tension becomes clear when we look at real‑world patterns documented in research and illustrated through composite examples.
 

Composite Example 1: A Criminal Justice Student Navigating Biased AI 

This example is a composite drawn from JFF’s findings on Black learners’ heavy use of AI for self‑directed study and the well‑documented racial bias in criminal justice algorithms. 

A young Black man studying criminal justice turned to AI tools to deepen his understanding of complex judicial concepts. These technologies helped break down case law and prepare for exams. Yet, a troubling pattern emerged. When he requested examples of policing scenarios, the system repeatedly generated situations centering on Black neighborhoods even without a specified location or demographic. He quickly realized that biased training data influenced these outcomes, imitating patterns seen in predictive policing tools. Such tools disproportionately target communities of color, as they often rely on historical arrest data rather than actual crime rates.
 

Instead of helping him understand the law, the tool was supporting the very assumptions he hoped to challenge as a future public defender. His experience demonstrates a broader truth: AI systems trained on biased criminal justice data can quietly reproduce the inequities they claim to solve.
 

Composite Example 2: A Black Journalist Confronting AI’s Blind Spots 

This example is a composite grounded in research on AI transcription errors, dialect misinterpretation, and the challenges journalists of color face in under‑resourced newsrooms.

​

A Black journalist at a legacy newsroom turned to AI transcription tools to speed up her reporting. Although the technology seemed promising, she encountered recurring difficulties: interview recordings with community members who spoke regional dialects and used culturally specific language were often misinterpreted. Hours were spent correcting errors her white colleagues rarely faced. AI-generated summaries also flattened the nuance of stories rooted in Black experiences, stripping away context essential for understanding the truth. 

These issues mirror findings from journalism researchers who have documented that AI transcription tools commonly struggle with African American English and other dialects, resulting in higher error rates in interviews conducted in communities of color. In a profession where accuracy is everything, these errors are not minor. They are structural. And they place an additional burden on journalists already operating in underfunded newsrooms with limited resources.
 

Why These Patterns Matter for AI Governance 

These composite examples reflect real concerns raised by the NAACP, which recently issued a comprehensive resolution calling for stronger protections across criminal justice, journalism, education, employment, and the creative economy. The organization argues that AI systems must be trained on diversified data, audited regularly, and governed by rules that prevent discrimination and misrepresentation (NAACP, 2024). Their resolution makes clear that AI is not only a technical issue. It is a civil rights issue.
 

In the criminal justice system, predictive policing tools have consistently sent officers disproportionately into Black neighborhoods. This does not occur because crime is higher; rather, it reflects decades of over-policing. Risk assessment algorithms, frequently used in sentencing and parole decisions, often label Black defendants as higher risk based on patterns unrelated to individual behavior. Without oversight, such systems can quietly reproduce the injustices they claim to solve.
 

Journalism faces its own AI‑driven challenges. Reporters from Black, Hispanic, Indigenous, and LGBTQ+ communities have described how AI tools often fail to recognize culturally specific language or situations (Radsch, 2024). Some newsrooms serving communities of color lack access to the latest AI tools, creating a digital divide within the profession itself. The NAACP resolution calls for protections for Black storytellers and creatives whose likenesses and voices are increasingly being replicated without their consent (NAACP, 2024). It also calls for education campaigns to help communities recognize misinformation and disinformation generated by AI, which can undermine political power and distort public understanding.
 

Concerns expressed in the JFF report highlight the immediacy of the issue. 71% of Black respondents reported needing new skills to keep pace with AI’s impact on work and education (Swartsel, 2025). Far from signaling fear, this response shows readiness. Black learners and workers are not waiting for the future; instead, they are already adapting. What they need now is a governance system that adapts to them.
 

Conclusion 

The award of equity in AI governance will not be given automatically. It must be earned through intentional design, inclusive policymaking, and persistent oversight. The stories of Black learners, workers, and journalists show what is at stake. They also show what is possible when AI is built with them rather than around them. To achieve this, stakeholders, including developers, policymakers, educators, and community leaders, must work together to ensure that AI systems are transparent, audited for bias, and shaped by the communities most affected. Concrete steps such as establishing oversight bodies, conducting regular audits, supporting culturally responsive data collection, and investing in community education should be central. If the world is serious about building fair AI systems, it must start by centering the people who have the most to gain and the most to lose. Equity is not a feature that can be added later. It is the foundation that determines whether AI becomes a tool of liberation or another chapter in a long history of exclusion.
 

References 

Alex Swartsel, A. (2025, September 30). Unlocking the promise of AI for Black Learners and workers. Jobs for the Future (JFF). 

https://www.jff.org/idea/unlocking-the-promise-of-ai-for-black-learners-and-workers/ NAACP. (2025, January 30). Ensuring representation and eliminating bias in artificial intelligence. NAACP. 

https://naacp.org/resources/ensuring-representation-and-eliminating-bias-artificial-int elligence#:~:text=BE%20IT%20FURTHER%20RESOLVED%2C%20that%20the%2 0NAACP%20will%20advocate%20for,particularly%20the%20African%20American %20community. 

Radsch, C. C., Nicol Turner Lee, I. P. H., Nicol Turner Lee, N. W., Keesha Middlemass, M. F., Fox, L., & Carmona, T. (2024, December 23). Journalism needs better representation to counter AI. Brookings. 

https://www.brookings.edu/articles/journalism-needs-better-representation-to-counter ai/