The AI Divide: What It Means and How We Fix It

Megan McDonnell
Accessibility
Design
Digital Divide
Fixes
Tenacity
UX
0 Comments
The AI Divide header image depicting a robot and a human building a bridge

You’ve probably seen the headlines. Generative AI (GenAI) is the fastest-adopted technology in history. After just two years, 39% of Americans are already using it, a rate double that of the internet at the same point in its life.

However, this incredible speed masks a deep and growing inequality. We’re at a critical fork in the road. This technology could become a great equalizer, offering a personal tutor to every student or breaking down language barriers for millions. Or it could become one of the most powerful engines of inequality we’ve seen in a generation.

For years, we’ve discussed the “digital divide” – the gap between those who have internet access and a computer, and those who don’t. Today, that gap is evolving into something new.

Welcome to the AI Divide. Two people can sit in the same computer lab, on the same internet connection, and still live on opposite sides of this divide. Though the problem is no longer just access, but also agency.

Going forward, we’ll explore this new chasm: what’s causing it, why it matters, and the practical steps we can all take to ensure AI becomes a bridge for everyone, not a barrier.

The New Gap: From “Digital Divide” to “AI Divide”

The old digital divide was about access to the “on-ramp” – getting a broadband connection or a smartphone. The new AI Divide is about what you can do after you’re online.

Experts sometimes call this the “second-level” or “third-generation” digital divide. It’s the gap between those who can simply access technology and those who have the skills to comprehend, control, and truly benefit from it.

The new inequality is between those who can effectively command AI to augment their work, creativity, and learning, and those who cannot. Now, the question is no longer just “Who is online?” but “Who can command AI?”

How GenAI Is Widening the Chasm: A Look into the New Divide

Without deliberate intervention, GenAI is on a path to exacerbate existing inequalities. It’s not just one thing; it’s a pile-up of four key factors.

  • The Cost of Entry: While many tools offer free tiers, they are often limited in their functionality. The most powerful models – the ones that provide a real advantage – require premium subscriptions that range from $20 to $200 per month. For a low-income household, that’s a significant barrier.
  • The Infrastructure Gap: GenAI tools demand high-speed internet and modern computers. Those requirements immediately exclude billions. Globally, 2.7 billion people remain completely offline. Even in the U.S., 24 million Americans lack high-speed internet, and the gap persists along racial lines: 83% of White adults have broadband, compared to just 68% of Black adults.
  • The Language Wall: Many leading models are overwhelmingly English-centric. That creates what some call a “monocultural AI landscape” reflecting a U.S.-centric “culture blob.” For non-English speakers, this isn’t just an inconvenience; it can mean 10x cost penalties and far worse performance.

2. The Skills Gap: A New Kind of Literacy

Using AI effectively isn’t just about typing a question. It requires a new “AI literacy” – the ability to ask the right questions (prompt engineering) and, just as importantly, to critically evaluate the answers you get back.

  • The Workplace Disconnect: A massive gap exists between what workers want and what companies are providing. While 94% of workers want AI training, only 5% of organizations are actually providing it at scale. Meanwhile, 66% of leaders say they wouldn’t hire someone without AI skills.
  • The Education Patchwork: Schools struggle to keep up. As of early 2025, only 28 states had published any AI guidance for K-12 schools. In higher education, 50% of colleges don’t give students institutional access to GenAI, often due to cost. The result is a literacy gap based on geography and wealth. For example, in 2024, 67% of low-poverty school districts offered AI training, compared to just 39% of high-poverty districts.

3. The Economic Divide: Who Profits and Who Is Displaced?

  • The “Winner-Takes-Most” Dynamic: GenAI is projected to create enormous economic value – between $2.6 and $4.4 trillion annually. But that value is heavily concentrated. In the U.S., over 50% of the economic gains are projected to go to the top income quintile, while less than 5% are projected to benefit the bottom quintile. Those with AI skills are already seeing salary premiums of 25–31% in key sectors.
  • Unequal Disruption: The flip side of this productivity is displacement. GenAI is projected to disrupt 300 million full-time jobs globally. That disruption doesn’t hit everyone equally. It disproportionately affects women and Black workers, who are overrepresented in the administrative and support roles most easily automated.

4. The Bias Amplifier: “Bias from the past leads to bias in the future”

Bias amplification is one of the most serious risks. GenAI models are trained on historical data from the internet. In the process, they absorb – and can actually amplify – the systemic biases found there.

  • Stark Racial Bias: The data is alarming. One study found AI resume-ranking systems preferred “White-associated” names 85% of the time. Commercial facial recognition systems have shown error rates of up to 35% for darker-skinned women, compared to less than 1% for lighter-skinned men.
  • Pervasive Gender Bias: There’s a gap in who uses the tech (men use it at rates 25% higher than women) and in how the tech sees women. Models have been caught assigning women to stereotyped roles like “domestic servant” while assigning men to high-status roles like “engineer.” The problem is made worse by the fact that only about 30% of people working in the AI field are women.

The Hope: Using GenAI as a Bridge for Inclusion

That was the bad news. But the picture isn’t all bleak.

The same technology that threatens to widen the AI Divide also has a powerful, unique ability to act as a democratizing force – if we deploy it intentionally.

1. Democratizing Skills and Knowledge

For the first time, the primary interface for a powerful technology isn’t complex code or menus; it’s natural language. That shift is a game-changer. It can empower individuals with low digital literacy to perform complex tasks, from writing a business plan to debugging code. Furthermore, it can also make vital information accessible by summarizing complex medical reports or legal documents into simple, easy-to-understand language.

2. Personalizing Education and Opportunity

GenAI has the potential to solve what educators call the “2 Sigma Problem” by providing a free, one-on-one “personal tutor” to any student with an internet connection, like the services offered by Khan Academy’s Khanmigo. It can act as a “College Coach” for first-generation students navigating the application process, or deliver lessons in multiple local languages, as seen in trials in India and Nigeria.

3. Overcoming Foundational Barriers

Here, the potential feels truly transformative.

  • Language: Real-time translation across hundreds of languages can empower non-English speakers in the global economy or help refugees communicate with doctors and aid workers.
  • Accessibility: GenAI offers life-changing tools for people with disabilities. That includes vastly improved screen readers, real-time captioning for conversations, and better speech recognition for atypical speech patterns, such as Google’s Project Euphonia.

The Action Plan: A Framework for an Equitable AI Future

So, how do we get the good scenario instead of the bad one?

The outcome is not guaranteed. It depends entirely on the deliberate choices we make right now. The work ahead isn’t just a job for coders; it requires a “whole-of-society” approach. Here’s a practical framework.

1. For Policymakers: Set the “Rules of the Road”

Policymakers can set the guardrails that determine whether AI widens or narrows inequality.

  • Invest in Infrastructure: Treat universal, high-speed internet as a foundational public utility, not a luxury.
  • Fund Public Literacy: Develop large-scale AI literacy programs in public schools, libraries, and community centers to foster critical thinking skills.
  • Mandate Accountability: Implement frameworks such as the Blueprint for an AI Bill of Rights. That means requiring transparency and independent audits for bias in any AI used in high-stakes areas such as hiring, healthcare, and criminal justice.

2. For the Tech Industry: Design for Inclusion, Not Just Profit

Technology companies decide who gets left behind – or brought along.

  • Prioritize Bias Mitigation: Bias can’t be an afterthought; it must be a core design principle. Invest in diverse, representative training data and actively engage marginalized communities in a co-design process.
  • Democratize Access: Continue offering powerful free tiers, provide subsidized access for schools and non-profits, and support open-source alternatives to prevent monopolies.

3. For Educators and Communities: Lead the Change on the Ground

Educators, nonprofits, and community leaders can turn AI from something done to people into something done with them.

  • Integrate AI Critically: We must move beyond simply banning AI in schools. Curricula should teach students how to use these tools responsibly, ethically, and effectively.
  • Empower “Digital Navigators”: Technology alone is never the answer. Support trusted “Digital Navigators” in local communities – real people who can provide personalized, culturally responsive training to those most at risk of being left behind.
  • Serve as Watchdogs: Civil society, researchers, and journalists must act as independent auditors to expose bias, hold industry and government accountable, and ensure the public good is being served.

Your Turn: The Choice We All Have to Make

GenAI holds up a mirror to our society. It has the potential to scale our best qualities – creativity, learning, and connection – or our worst – bias, inequality, and exclusion.

The goal can no longer be just access to AI. The goal must be equitable outcomes. We have to shift our focus from “Who can use this tool?” to “Who benefits from it?”

The path forward must be human-centered. The choices we make in the next few years will decide whether GenAI becomes a tool that serves all of humanity or the most powerful engine for inequality we have ever built.

Here’s one small action you can take today: start a conversation. At your next team meeting, community gathering, or even at the dinner table, ask: “How are we using (or planning to use) AI? And who might be getting left out of this conversation?”

At Tenacity, we believe building inclusive, accessible, and ethical technology isn’t a feature; it’s the foundation. If you’re wondering how to make your website, digital products, or AI strategies more equitable, let’s talk!

 Sources:

Tags:
Facebook X / Twitter LinkedIn

Leave a Comment.

5 × two =