What the U.S. is doing in AI literacy

🏫 Schools

• National AI challenge for elementary and high schools and teachers. Students propose or build an AI solution to a problem in their own community. It has state, regional, and national stages.
• Company-led teacher training: Anthropic’s open curriculum (Creative Commons), US$ 150 mi from Google, 10,000 educators through Amazon by 2028, and more than US$ 20 mi from Meta.
• Education grants start giving preference to projects with AI.

👷 Workers

• Official framework AI literacy framework, detailed below.
• Job retraining funding can pay for AI courses for unemployed and displaced workers.
• Professional learning: a goal of 1 million learners, with funding distributed to every state.

🏪 Small businesses and communities

• AI-Ready America: AI literacy and practical skills, technical assistance for small businesses and municipalities adopting AI, and internships and real-world projects.
• One hub per state: US$ 1 mi per year for 3 years (plus one possible additional year), 56 hubs in total.
• Includes the rural extension, bringing AI literacy to the field.

🏆 The challenge in numbers

• More than 20,000 students from all 50 states, the capital, Puerto Rico, and 49 overseas military schools.
• Award-winning projects: homework help, bullying prevention, support for people with visual impairments. The award-winning teachers teach AI within the subjects they already teach.
• The ceremony took place amid a growing backlash against screens and AI in schools.

🗺️ The states

• 77 bills in 27 states about AI in the classroom only in 2026.
• Examples: Maryland (AI policy in each school district and AI literacy), Idaho (literacy standards, teacher training, AI does not replace teachers), Oklahoma (AI only with teacher supervision).
• California already requires AI literacy to be included in math, science, and history curricula.

🏛️ Congress and government

• No federal law AI literacy framework approved by September 2026. There are projects for schools and small businesses; the latter have already been approved by the House.
• New Department of Labor funding to retrain workers affected by AI (Jun 2026).
• The government trains its own employees: an official 2026 AI course and training with Stanford and Princeton.

📊 What the research shows

• 6 out of every 10 teachers used AI during the 2024–25 school year; 3 in 10 use it every week and save about 6 hours per week (Gallup, Jun. 2025).
• 26% of teenagers use ChatGPT for school assignments, twice as many as in 2023 (Pew, Jan 2025).

✅ What’s real

• There is a decree, funding, deadlines, and a published framework.
• The national challenge actually took place, with the final in June 2026.
• The framework is good: practical, tied to professions, and focused on evaluating responses.
• The focus on small businesses and rural areas reaches people who are usually left out.

⚠️ What raises concerns

• Patchwork: the programs are scattered, and the results depend on each state taking action.
• Small amount of funding for a country this size. US$ 1 mi per year per state is too little.
• Big tech companies set the direction: companies train teachers using their own tools. Who decides how AI enters the classroom?
• Few safeguards privacy and outcome evaluation in the grant programs.
• The Department of Education is being dismantled, with the risk of losing the people who know how to run the programs.
• AI literacy without digital literacy: cuts to digital inclusion funding affect the local organizations that would provide the foundation (Brookings).
• The stated motivation is the geopolitical rivalry. Literacy is treated as part of a race.

1. Course checklist

Use the 5 areas as a checklist: does every AI course cover all five? Areas 4 (evaluating responses) and 5 (using AI responsibly) are the ones most often missing.

2. Teach through the profession

AI for lawyers, accountants, healthcare professionals, and store owners, each with their own examples, teaches more than a generic lesson.

3. Community challenge

The challenge format (solving a real problem in your neighborhood or city with AI, in stages) works in a school, group, or online community.

4. Open material

Open-license AI curricula already exist that can be adapted and translated for Brazilian teachers.

5. Small business first

In the U.S., small businesses and local governments have their own dedicated initiatives. In Brazil, these are the groups that need the most support and receive the least.

6. Neutrality

Teaching without relying on a single provider is an advantage over training funded by big tech companies.