AI, Equity, and Agency: The 3 Big Areas Education Leaders Are Focused on Right Now

At InstructureCon 2026, more than 100 of the 275 breakout sessions touched artificial intelligence. That number might have been easy for most people to predict. What’s maybe more interesting than a straight number session count on a topic, though, is what sat next to AI on the agenda. Session after session paired it with two concerns that have lived on the desks of superintendents and provosts alike for years: reaching every learner, and helping learners take ownership of their own progress.

The same pairing showed up at ISTELive 26, where ISTE built a pavilion around a Portrait of an AI-Ready Graduate, named "Ignite Agency" as a core learning principle, and gave equity its own track under "Opportunity, Inclusivity, and Cultural Competency." ISTE's program focuses on K–12, but the three themes really matter for educators everywhere because they all address preparing every student for life outside the classroom. 

So whether you run a district or a campus, the conference programs make a strong case for three areas that deserve your planning time this year: AI, equity, and agency. None of them is a decision you make once. Each one requires the judgment to weigh options well, the flexibility to adjust as conditions change, and the footing to act with confidence when the evidence is still coming in. 

Here's why each one matters across K–12 and higher ed, and why it might influence high-level academic strategy.

The focus on AI centers on responsibility

Two years ago, AI sessions were mostly demos, focused mostly on “Look what this tool can do.” The rooms that filled up in 2026 were about judgment: how to teach students to check AI output, how to design an assignment that decides where AI belongs, and how to build AI literacy into what a teacher is expected to know. One session, "From Pilots to Proof: AI That Boosted College Admissions Success," showed the same shift playing out on the higher ed side, where the interest has moved from pilots to evidence of what actually worked.

The pressure looks a little different depending on where you sit. In K–12, one ISTE session ("Keeping Gen Alpha Human") reported that 86% of children ages 9 to 17 already use AI, according to research from Dr. Yi Zhang. In higher ed, it’s more about protecting academic integrity and what a degree really certifies when a chatbot can draft the essay. Students are going to use AI, but what do educators believe that AI should never replace?

Dr. Zhang's framework identifies four capacities worth protecting: the ability to self-regulate, to listen inward, to build identity beyond achievement, and to relate to others. Richard Culatta, CEO of ISTE+ASCD, made a parallel point from the main stage, arguing that the next chapter is using AI to strengthen human skills rather than substitute for them. 

Culatta framed it on the Edtech Insiders podcast this way: schools should model "using AI to help students be better at being human." Instructure's position lines up with that thinking. AI works best when it's built into the tools educators already use and pointed at instruction, not bolted on as a separate destination. IgniteAI, the AI built into Canvas by Instructure, follows that principle by keeping the instructor in the lead rather than just in the loop on both K-12 and higher-ed campuses. 

The strategic implication is the same at either level: write policy around purpose before you write it around tools. Name the friction you're trying to relieve, whether that's faculty capacity, feedback quality, or integrity, then judge vendors against that goal.

Equity is now an access-and-design question

Equity showed up in roughly 48 InstructureCon sessions, with fewer sessions treating it as a standalone initiative. More treated it as a design requirement built into everything else. Titles like "Crafting Stories for All: A UDL Pathway to Engagement" show equity moving into instructional design and universal design for learning; that language travels well into higher ed, where completion and persistence gaps are versions of that same equity and accessibility story.

At ISTE, Gregory Trieste's session on neurodiversity landed on a line worth repeating in any leadership meeting: if you're in education, you're in special education. In a district, that means serving more students in general education settings. On a campus, it means disability services and accommodations can't be a separate office that content routes around. Instead of asking which learners receive accommodations, more educators are considering whether the core tools and materials work for everyone by default, including first-generation students, multilingual learners, and students with disabilities.

This is where AI and equity meet, and where it’s important to consider the potential risks. A session titled "Mirror or Mirage? What AI Images Reveal About Hidden Bias" notes that the same tools that can personalize learning can also encode bias at a speed and volume no human review can catch. For strategy, equity can't be a downstream cleanup step. It belongs in procurement questions, in how you test tools for bias and developmental fit, and in whether accessibility is a real requirement or a checkbox.

Agency is the outcome

Agency, student voice, and ownership appeared in more than 30 InstructureCon sessions, from "Choice Boards That Drive Student Agency" to "From Consumers to Creators: Amplifying Student Voice with Design Thinking." In higher ed, we call it self-direction, and it's a key predictor of persistence and completion.

One ISTE session, "Beyond Engagement: How AI and Hope Ignite True Student Agency” explained that hope is a psychological state that predicts achievement and resilience, and it can be taught. Their claim: the best use of AI helps a learner find another route when they get stuck, rather than handing over the answer. Because, as in all things AI in education, there is caution. A separate ISTE session on learned helplessness listed over-scaffolding and over-prompting among its root causes. AI that does the thinking for a learner produces the opposite of agency. So the design goal is narrow: tools should widen the paths a learner can take toward a goal, not shorten the distance to a finished product.

What this means for your strategy

The conference programs are a mirror that reflects what thousands of educators chose to spend their limited time learning about. What they show is a field, from kindergarten through college, that’s moved beyond asking whether AI technology belongs in education to a harder (and better) question: how do we make it work for every student, and how can a growing focus on AI, equity, and agency help prepare students for lifelong learning and life beyond the classroom?

Leaders who work these three areas together end up building something none of them delivers alone: an institution that can take in new information and change course without losing its footing. For leaders, there are a few edtech-specific steps to support the shift. 

• Define the human capacities you want to protect before adopting any AI tool, and write them into your evaluation rubric. 
• Make accessibility and bias-testing part of procurement, not a review you run after the contract is signed. 
• Judge every tool by one question that cuts across all three areas and both segments: Does this help a learner do something they couldn't do before, or does it just do it for them?

Check out the resources below to get more insight into AI, equity, and agency for learners across K12 and higher education.