By Cailin Acosta in La Mesa, California
Artificial intelligence is becoming part of everyday life in classrooms, changing how educators prepare lessons, how students conduct research and how information is shared.
But when the subject is the Holocaust, the opportunities presented by AI come with a particularly important responsibility: preserving historical accuracy and the authentic voices of those who experienced it.
That is one of the questions addressed by ADL Ora, a closed artificial-intelligence assistant developed by the Anti-Defamation League specifically for K–12 educators.
Launched in February 2026, ADL Ora is designed to help educators teach about Jewish identity, antisemitism and the Holocaust. The tool draws from ADL Education materials, including Echoes & Reflections, rather than searching the open internet for information. According to the ADL, Ora is intended to help educators plan instruction, navigate difficult student questions, respond to antisemitism in school settings and adapt resources to classroom needs.
I recently completed the ADL’s Teaching with AI: Antisemitism, Jewish Identity & the Holocaust course and had an opportunity to explore Ora for myself while completing education requirements for my Social Work license.
The experience left me thinking less about whether AI belongs in Holocaust education and more about how educators can use it responsibly without allowing technology to replace historical sources, professional judgment or the human stories at the center of Holocaust education.
As part of the course, I asked Ora to help me prepare a 30-minute presentation for students in grades 9–12 following an incident involving antisemitic graffiti in a school parking lot. I explained that the students had very little prior knowledge. I then revised my prompt to ask for inclusive language so students with learning, language-based and cognitive disabilities could participate in the discussion.
I found the response helpful. I also entered the same prompt into ChatGPT, which produced some similar recommendations. The comparison was interesting. ChatGPT provided a more detailed presentation, including a breakdown of individual slides, while Ora produced more of a shorter guideline that left room for flexibility. I could have continued refining the Ora prompt by asking specifically for a slide-by-slide presentation.
Both responses also addressed an important question: whether showing antisemitic imagery provides enough educational benefit to justify the potential harm of displaying it to students. For me, this was a reminder that the technology can help generate ideas, but the educator still has to make the final decision about what students should see and how a difficult subject should be introduced.
AI is creating new possibilities for Holocaust scholarship and remembrance. The ADL and Echoes & Reflections course highlighted several examples of technology being used to make historical records more accessible, including projects involving photographs, archival records, survivor testimony and victim identification. AI and related digital technologies can help researchers process enormous collections of historical documents and make information more searchable. There is also a growing effort to preserve survivor testimony in ways that future generations can access.
The potential is significant: Technology may help researchers discover information that has remained difficult to locate and help students encounter authentic historical records that otherwise might be inaccessible to them. But the technology must remain connected to the original sources.
One of the most important ideas from the course was the distinction between what AI can assist with and what educators must continue to decide themselves. AI can help an educator organize a lesson, develop discussion-question stems or differentiate material for different reading levels. But educators must verify historical accuracy, select appropriate primary sources, determine what is emotionally and developmentally appropriate, and decide how to facilitate difficult conversations.
The course also establishes several boundaries for Holocaust education. Educators should not ask AI to role-play as a Holocaust victim, survivor or perpetrator. Nor should they ask AI to recreate atrocities that were not photographed or filmed. AI-generated historical images can be mistaken for authentic primary sources. Similarly, AI-generated quotations attributed to real historical figures should never be accepted without verification against reliable archives or collections. The distinction is simple but important: AI can help educators work with historical material; it should not manufacture the history.
One of the more interesting ideas presented in the course was to use AI’s limitations as an opportunity for teaching historical thinking. Students could ask an AI system to produce a biography of a lesser-known Holocaust figure and then compare that response with information from a trusted historical archive. Instead of simply asking whether the AI answer is “right” or “wrong,” students can examine what it invented, what it omitted and how they can verify the information. That activity teaches skills historians use every day: sourcing, corroboration and contextualization. It also teaches students an increasingly important lesson in the digital age: a confident answer is not necessarily an accurate answer.
The conversation has particular relevance for California educators. California requires Holocaust education in grades 7–12, making questions about appropriate resources and instructional practices especially relevant to middle and high school classrooms. For educators working within that framework, tools such as Ora may provide another way to develop lessons and prepare for classroom discussions, while still requiring teachers to consult the underlying educational and historical sources.
My overall experience with ADL Ora was positive. I am interested in continuing to compare Ora with other AI platforms. In my own practice, I could see returning to Ora for work involving bias-based harm, antisemitism and Jewish education. I also appreciated that the course introduced me to Echoes & Reflections, and I subsequently signed up for its newsletter.
Most importantly, the course reinforced something I already believed: AI is not going away, so educators need to understand how to use it thoughtfully. I came to the course because I am concerned about preserving the memories and stories of those who survived the Holocaust. I left thinking that AI can be part of that work, but only if we remain careful about what we ask it to do and equally careful about what we refuse to let it replace.
*
Cailin Acosta is the assistant editor of the San Diego Jewish World. This article reflects her experience completing ADL’s Teaching with AI: Antisemitism, Jewish Identity & the Holocaust course and exploring ADL Ora.