What AI in Schools Looks Like From the Inside
Joe Eckstein has spent years at the intersection of data protection and education technology. Here's what he sees...
What do the people who actually live inside school systems—managing technology decisions day after day—think about AI?
About four years ago, I joined a data privacy advisory group supporting the New York State Education Department. The committee draws from a range of backgrounds, but its backbone is the Data Protection Officers (DPOs) embedded in local school districts across the state. These are the people who deal with the tough, consequential work of protecting student data before it ever becomes a headline.
One of these leaders has become a friend. Joe Eckstein has had the kind of career (spanning both industry and education) that gives him a rare insider/outsider lens on how schools actually navigate technology, as opposed to how policymakers imagine they do. That perspective has made him one of the more clear-eyed thinkers I know on data privacy—and someone who deeply understands what AI actually means for schools and society.
Three years ago, Joe and I collaborated on an AI-related project for the state’s then Chief Privacy Officer that, in retrospect, was asking all the questions we’re now scrambling to answer. One of the things we kept returning to: the core technology infrastructure has to be solid in school districts before you can meaningfully grapple with AI’s particular challenges.
Without policies and controls that can hold up to any technology, layering on new ones just compounds the problem. Joe sees where the foundation is missing. That’s exactly why I wanted his perspective to share with families...
Meet Joe Eckstein
What is your current role?
I am a Business Manager running the contracts department at Eastern Suffolk BOCES and I also serve as the agency Data Protection Officer (DPO).
What has your career journey looked like?
After law school, I worked as a litigation attorney handling contract disputes. After 5 years of litigation I took more of an interest in operations management and by 1998, I left litigation and I took a position as an administrative law judge with a NYC government agency. In 2000, the NYC agency where I was working as a judge convinced me to take a supervisory role in their general counsel’s office.
Then 9-11 happened and that changed a lot for me. I quickly climbed the ladder to become a Deputy Commissioner and was running special projects as well as managing a staff of over 200 people. In 2008, I left public service to return to litigation. In 2013, I opened my own 40 person law firm working as in-house counsel to an insurance company.
In 2015, some personal decisions caused me to take sabbatical for 2 years and when I decided to return to work, I chose to return to public service and I took my current position with Eastern Suffolk BOCES. There are common threads in my career of always staying involved in technology and legal ethics. Becoming a DPO was a natural progression for me.
As a Data Protection Officer, what do you wish more stakeholders within a school better understood?
I am often surprised by how few people think sharing data is a potential safety issue for children. There is an assumption that all children are actively sharing their data via social media, video games and apps and that keeping school data safeguarded is a futile endeavor in light of how children behave online on their personal devices.
I wish more stakeholders understood that securing student data is not just a matter of technical compliance with a frivolous rule but that children’s safety and identity are at risk. This is especially true of those pieces of data which are easily used to steal someone’s identity, and also geo-location data.
What do you feel is the biggest risk to data privacy today?
In the school setting, teacher and administrator ambition to provide cutting edge technology to children. This happens too often even if the tech provider is unwilling to provide privacy safeguards. Administrators need to just say no to private tech companies.
In a more broad societal view of the question, the biggest risk is peer pressure. Everyone wants to be a part of the social media and AI trends. Consumers are increasingly being convinced that these tech products will serve individuals best the more that we share information about ourselves. There is truth in that; the more personalized the product the more useful it may be to individual consumers. But it is also true that it is the biggest risk to privacy. It’s hard to claw back privacy once you’ve given it up.
What do you think about AI both personally and professionally?
I think specialized AI is amazing. What it can accomplish in terms of consolidating so much information into condensed usable forms is truly remarkable. Nevertheless, I have concerns about lack of regulation and the lack of understanding of the way in which this technology can encroach on human autonomy.
As an attorney, I have real concerns about the usurpation of copyrighted material in the training models. As a parent and a human, I have concerns that the mental health impacts of social media are growing and will continue to grow exponentially as humans rely more on AI technology for companionship and rely less on human-to-human contact.
I am very concerned about the displacement of humans from the workforce, especially as LLMs are moved away from human operated computers and become part of autonomously operated robots and machines. At its current trajectory, I am concerned that AI will outgrow humans and topple democracy—a concern, I might add, that is shared by many executives and scientists in the field.
As a DPO and contracts manager, I have more specific concerns about the legal responsibility of the tech vendor vis-a-via the potential mishandling of data by an AI subprocessor. Once you remove the human from the equation, what incentive is left to follow laws written for humans?
My focus is on ensuring the language in the contract clearly defines who is responsible for safeguarding the data and what remedies are available to an aggrieved party. I am always looking to see if the humans are trying to pass on the responsibility to the consumer or the digital AI entity; an entity which cannot be made to account for its’ wrongdoings. Many service as a software providers have quietly added AI to their back end analytics and are simply keeping the old contract language in place. This is dangerous in terms of data privacy. All parties to the contract need to be transparent about where data is going.
What are you most optimistic and pessimistic about when it comes to the future of technology?
A big problem that has plagued K-12 education for over a century is the divergence of learning readiness among the students in a given classroom. AI will, for the first time, allow real time assessments of a students’ readiness to learn any given topic and allow students to virtually join classes which are targeted for any specific audience.
Geographic barriers will disappear and each day, students will find themselves with new classmates all at the exact same learning point. It will eliminate students who are lost due to being behind in the learning or bored because they are way ahead of their classmates and the teacher. I am not sure how long it will be before something like this is available given the power of the teacher’s unions—but it is coming; someday.
I am most pessimistic about humans maintaining control once robots with AI powered “brains” enter the workforce. This will cause a societal disruption that will change the course of education forever. Removing the classroom from the equation solves the problem of readiness to learn—but removing the human from the learning process is a step in the wrong direction. It will have social consequences that will hinder the child’s development. Social media is already having this impact on select populations—AI will exacerbate that problem.
Do you know who manages data privacy in your school? It’s worth reaching out to them. We all need to work together as stakeholders to protect the infrastructure that AI will increasingly disrupt. Have a question for Joe? Let me know!