OpenAI committed $150 million in grants over three years for digital wellbeing, teacher training, and AI education. IBM pledged to skill two million learners by 2028. Code.org will engage 25 million learners in an "Hour of AI." Google.org gave $10 million to the Manufacturing Institute for AI training on factory floors.
These are serious numbers. And they raise a serious question. If this much money is flowing into AI literacy, why does the average classroom teacher still have nothing practical to use on Monday morning?
Where the money actually goes
OpenAI deployed "ChatGPT for Teachers" across districts covering over 150,000 educators. They launched the OpenAI Academy with free training content for community college faculty. They ran pilot programmes in countries from Estonia to Trinidad and Tobago.
This is product distribution dressed as education. That is not a criticism. OpenAI has a legitimate interest in teachers knowing how to use its product well. But there is a difference between training someone to use ChatGPT and training someone to teach AI literacy.
A History teacher who completes an OpenAI training module knows how to use ChatGPT in the classroom. She does not necessarily know how to connect the concept of algorithmic bias to source evaluation. She does not know how to explain to Year 9 why an AI-generated timeline might omit inconvenient facts. She does not have four activities ready to go that connect her lesson on the Cold War to the way AI systems handle contested narratives.
Product training teaches the tool. Literacy teaching builds the understanding.
The Digital Promise and TNTP model is closer
In April 2026, Digital Promise and TNTP announced a three-year collaboration funded by Salesforce. Their goal is to reach up to 15 million students by 2028 by testing AI implementation strategies in real classrooms, grounding them in research, and producing district-level guidance.
That model is closer to what schools actually need. It starts in the classroom, not in a product demo. It produces replicable guidance, not brand-specific training. And it acknowledges a reality that the press releases from tech giants tend to skip: public school districts generally lack the internal R&D capital to independently develop research-backed playbooks for AI integration.
Schools are not equipped to build this from scratch. They need it built for them.
NVIDIA is doing something different
NVIDIA signed a deal in Utah to certify teachers through its Deep Learning Institute University Ambassador Programme. Public school teachers get access to teaching kits and GPU-accelerated cloud workstations. In California, NVIDIA partnered with a city government to connect high school robotics pipelines to higher education and municipal infrastructure.
This is pipeline thinking. NVIDIA is not training teachers to sell its products. It is investing in teachers because teachers are the multiplier. One trained teacher reaches 150 students a year. Train 100 teachers and you reach 15,000 students. The economics of teacher training are better than any ad campaign.
But even NVIDIA's model has a gap. It focuses on Computing and STEM. AI literacy does not live in one department. It lives across the curriculum. A PE teacher needs to understand algorithmic decision-making in fitness tracking. An Art teacher needs to engage with authorship questions around AI-generated images. A Business Studies teacher needs to examine how companies use AI for hiring decisions.
None of those teachers are going to attend a Deep Learning Institute programme. They need something that meets them in their subject, in their lesson, on their terms.
The gap in the market is the gap in the classroom
$150 million from OpenAI. $750 million from Google. $10 million from Google.org. Hundreds of millions more from Cognizant, IBM, Salesforce, and NVIDIA.
All of that money, and the Geography teacher in Coventry still does not know how to connect her lesson on urbanisation to AI literacy.
The money is going to platforms, partnerships, certifications, and academies. It is not going to the individual teacher who needs four activities and a coaching script by period 3 on Tuesday.
That is the gap AILitKit fills. Not a platform. Not a certification. Not a 12-week course. A tool that takes the lesson you are already teaching and shows you where AI literacy lives inside it. 1 minute. Any subject. Coaching notes so you know what to say. Support, challenge and differentiation so every student can access it.
The billions are flowing. But until someone solves the problem at the lesson level, most teachers will keep doing what they have always done: nothing, because nobody showed them how.
Matthew Wemyss is the founder of AILitKit and IN&ED, and author of AI in Education: An Educator's Handbook.