Cultivating student agency with Amazon’s Danielle Perszyk
Amazon AGI SF Lab cognitive scientist Danielle Perszyk took the stage for the second keynote of the day, challenging attendees to ask a deeper question: Will AI enhance or erode our agency as learners, educators and institutions?
“Agency is about having a mechanism to learn and act, understanding the focus, and habitually thinking rather than reacting,” shared Perszyk.
What makes agentic AI transformative, Perszyk explained, isn’t just its ability to automate tasks, but its potential to redefine what learning looks and feels like: adaptive, personalized and grounded in each learner’s intrinsic motivation.
It’s not just about having AI tutors; students also need a multi-agent system of universal teams to cultivate agency: a team of agents to help them learn. Perszyk explains that sometimes students use chatbots without actually learning — and we need to fix this. “Students need to know how to ask the right questions, check and critique sources,” she said.
By implementing a multi-agent system, the lines of learning and teaching become blurred and true learner agency is achieved. “This gives us the power to learn anything we want to become true lifelong learners,” Perszyk said.
For institutions to be truly effective in this implementation, data will be central. “We need the right types of redundant data. Intelligence doesn’t live in an individual, but in groups and how different groups of students learn.”
In the era of agents, we need to be lawful that they give us more agency than take it away. This means creating collective agency; to do so, Perszyk explains, we need to coordinate our efforts in tech and higher education.
New global collaboration: AWS, ASU and Cintana Education
Perszyk welcomed Colleen Schwab, John Rome, and Emiliano Diez onstage to announce a novel collaboration between AWS, ASU, and Cintana Education to bring agentic AI-powered solutions that reshape how AI can advance higher education worldwide.
“One of the most powerful assets in this collaboration is our ability to test and refine AI solutions across a truly global learning ecosystem,” said Diez, Chief Technology Officer at Cintana Education.
Together, the three organizations will harness AWS cloud technologies, ASU’s academic expertise and Cintana’s global university network to co-create solutions that directly serve students.
“Few in higher education have access to such diversity of institutions, student populations, and local challenges,” Diez shared. “This live, multinational testbed allows us to accelerate learning cycles, adapt tools for real-world use, and ensure that innovation is not just scalable — but inclusive and relevant across geographies.”
Students are at the center of this project, with student developers working under the guidance of cloud experts from AWS and ASU’s AI Cloud Innovation Center.
The first solution in development is an agentic AI-powered support agent that engages prospective students during their admission journey, enhancing efficiency in student recruitment and support.
"As a student myself, I’ve felt very connected to this project from the start," said Pathak, who contributed to the project as a student worker. "We're designing the AI system to walk alongside students through their application process."