Pearson’s Dave Treat: AI Should Guide Learning, Not Cheat

Pearson’s Chief Technology Officer, Dave Treat

On academic integrity, AI adoption, and the critical skills that matter as agentic AI transforms education

Dave Treat joined Pearson nearly two years ago as Chief Technology Officer, later assuming the role of Chief Information Officer, thereby overseeing the company’s entire technology portfolio.

This move marked a return to the sector he grew up in.

“My favorite part of this job is returning to the world of education,” he says. “Education was my original career path. Everyone else in my family is a teacher in one way or another. Now, I can hold my head high knowing that I too am helping people realize the life they imagine through learning.”

The role spans strategic innovation and keeping a 182-year-old company running day to day.

“I am equally responsible for our agentic AI future as I am for fixing the space bar on our CEO’s laptop when it breaks,” Dave explains. “My heart is in innovation, but I also need to keep this place running.”

It is certainly no small task, considering Dave’s technical oversight covers a global workforce of approximately 17,000 people across multiple business divisions.

Pearson’s Chief Technology Officer, Dave Treat. Credit: Pearson

Academic integrity concerns are real

One of my first questions for Dave was how Pearson addresses cheating as students gain access to generative AI tools.

Dave is clear that the risk is genuine, but argues it stems from how the tools are deployed rather than from the technology itself.

“It’s important to recognize that the academic integrity concerns are real,” he says. “AI is a real concern, but only if we just dump it on teachers and students. If they only use it to get answers, it leads to cheating and stops actual learning.”

Pearson’s AI study tools instead prompt students to think through problems rather than hand over answers, guided by learning science.

Dave notes that usage data reveals a pattern that has little to do with classroom hours.

“We’re getting the peak hit rate of US students around 9pm in higher education,” he says.

“So, they’re asking a question definitionally when the teachers and TAs are not there to help them.”

The same interactions feedback to teachers the next morning, giving them a sense of where the class is struggling before the next day of tutoring begins.

Did you know?

• 9pm is the peak hit-rate window when US higher education students are most actively using Pearson’s AI study tools

Why AI needs a driving school

Asked why so many organizations invest in AI technology without first preparing their workforce to use it, Dave points to a pattern he has observed across several technology cycles.

“Having been in the world of technology now for more decades than I want to count, this is not a new pattern,” he acknowledges. “We’re all subject to the next hot tool and want to apply it.”

Dave uses the analogy of his 16-year-old son learning to drive to explain his approach to investing in new technology.

He argues that throwing powerful technology at people without guidance is like handing a teenager car keys without lessons – it inevitably leads to a wreck. Instead, he emphasizes the need for structural training, similar to driving school.

A central team for consistent AI standards

Pearson’s response to that challenge has been to build an AI Centre of Enablement, spanning customer-facing products, engineering and internal operations.

From the outset, Dave’s team chose to dive straight into the technology, focusing holistically on how AI could transform the business across three key dimensions.

To keep the company aligned, a small central team identifies what is working in specific areas and shares those successes more widely.

“There’s a ‘find good and diffuse’ aspect to it,” Dave says. “We knew right away that we would need to apply common standards and protocols.”

Underpinning that effort is a wider modernization of Pearson’s data infrastructure, which Dave admits is still catching up with the ambitions of the AI program.

“We have quite a way to go on our data foundations as I think most companies do,” he says. “We are a 182-year-old company and it’s fair to say we don’t have 182-year-old tech! But we’re always looking at how we can modernize and enhance it.”

Collaborating with big tech

Pearson works with several major technology companies that also build their own training and learning tools, a dynamic some might frame as competitive.

However, Dave perceives it differently: “We don’t think of it as competition in any way. “We’re actually partnering with many of them and we’re moving to being in a much more open ecosystem.”

He describes the arrangement as mutually reinforcing, combining Pearson’s education content and learning science with partners’ platforms and distribution.

“Our partners have some fantastic chief learning officers and skills and capabilities, and so we’re just here to help amplify them and vice versa.”

Asked to describe a university student or corporate learner in 2030, Dave returns to a common theme: meeting people within their existing routines rather than pulling them away. Credit: Getty Images

Judgment and imagination remain essential skills

For organizations outside the technology sector adopting AI, Dave argues the underlying skills required have not changed much across previous innovation waves.

“It always comes down to the quality of a person’s judgment, their domain expertise and their imagination,” Dave says. “The more depth they have in their field, the more imaginative they can be with it.

“Combine that with empathy and collaboration, and those are the human characteristics that matter. No matter the technological wave, if you have those skills, you can make the most out of any technology – and AI is no different.”

Key fact

• Pearson has an approximate 17,000-strong global workforce under Dave’s technical oversight

Learning in the flow will define the future

Asked to describe a university student or corporate learner in 2030, Dave returns to a common theme: meeting people within their existing routines rather than pulling them away.

“What actually works is meeting people where they are and injecting microlearning experiences into their daily lives, because we now have visibility into their specific context,” adds Dave.

“The future of learning is in the flow – in the flow of life, the flow of work and the flow of education.”

That thinking extends to assessment, too. Dave suggests the oral exam, long recognized as a strong indicator of understanding, may finally become practical at scale through AI-guided interaction grounded in a professor’s own material.

“We’ve known for a very long time that oral exams provide a high-quality signal of a student’s understanding of a particular topic,” Dave goes on. “A student sitting in front of a professor, needing to articulate exactly what they know, truly demonstrates their depth of learning.”

For organizations outside the technology sector adopting AI, Dave argues the underlying skills required have not changed much across previous innovation waves. Credit: Getty

I am as equally responsible for our agentic AI future as I am for fixing the space bar on our CEO’s laptop when it breaks

Dave Treat, CTO, Pearson

Dave is equally direct about where he thinks the industry gets ahead of itself. He describes talk of fully autonomous agents replacing jobs as overstated, while pointing to agent verification – the process of assessing what a given AI agent is actually capable of – as an area deserving closer attention.

Pearson has already built a working AI Agent Verification prototype in that space, drawing on its background in assessing people rather than machines.

Built with IBM Consulting, the prototype applies the company’s decades of human credentialing experience to autonomous machines. The system continuously tests and audits digital agents to certify they can safely execute corporate workflows without veering outside their designated boundaries.

By treating software agents like employees who need to earn a skill credential, the platform provides businesses with a standardized framework to govern and trust their AI workforce.

Reflecting on what continues to motivate him, Dave returns to something more personal than any product roadmap.

“It takes a lot of courage to learn,” he says. “Learning, actually, is meant to be hard. If it’s not difficult and you’re not getting your brain to actually do work, then you won’t learn.”