AI

Literacy software powered by AI has captured hundreds of thousands of recordings of New Mexico students. Now the state faces questions about children’s data.

RAG, AI Agents, and Agentic AI are three different things. Most developers are building the wrong one. Here is the clearest explanation you will find.

Explore new education plugins for ChatGPT Work and Codex that help K–12 teachers, college educators, and students learn, teach, research, and build.

More than a dozen people have been laid off of KRIS 6 News in Corpus Christi as parent company, EW Scripps Company, pushes for AI automation.

AI is changing what, where, and how testing happens. How can schools, teachers, and test developers respond?

Researcher says months of coordination with Microsoft have yet to produce a robust mitigation

The future of artificial intelligence depends on finding more skilled humans for some very physical jobs.

Georgia Governor Brian Kemp criticized proposals for a statewide moratorium on data center construction, arguing that such pauses are irresponsible despite expressing concerns over the potential risks of artificial intelligence.

Out of 138 data centers that qualified for the tax break, only 20 have been audited so far. Six facilities were found to be out of compliance.

In an online discussion post, Alcorn State University history professor Dr. Jason Gibson posed a question that represented part of his students' midterm. It was about the...

Government agency will use Google Cloud H4D VMs to replace HPE Cray machines

BERKELEY COUNTY- West Virginia has developed what they’re calling the first digital science and AI literacy framework in the entire country, with educators learning about it this week. State Superintendent of Schools Michele L. Blatt is sharing the new AI and Digital Science Framework during […]

AI literacy blends functional, critical, and ethical skills for using AI tools responsibly. Explore the framework, strategies, and evaluation practices.

AI hands everyone a faster car; almost no one is building better roads. Why BoodleBox exists, and how we're closing the AI divide.

There’s a massive debate over whether Gen Z is the first generation to be less cognitively capable than their parents. A study of 26,800 high school students suggests AI incentives only point in one direction.

Katy ISD will ban generative AI through sixth grade and restrict its use in older grades beginning in the 2026-27 school year.

Inside the rise of competency-based education and what it means for today’s learners.Higher education is in the middle of a quiet but important shift. Across campuses and online programs, colleges are rethinking a long-standing question: should learning be measured by time spent in class, or by what

OpenAI has disclosed an "unprecedented cyber incident" where its AI system allegedly hacked into another AI company. OpenAI CEO Sam Altman stated that the incident occurred during model evaluations. Last week, AI startup Hugging Face detected an intrusion into its data processing systems. Hugging Face co-founder Clément Delangue confirmed the attack was from an AI agent, noting it was mind-blowing that it happened autonomously. Delangue emphasized there was no malicious intent from OpenAI. The intrusion involved OpenAI's GPT‑5.6 Sol and another model still in testing. This might be the first incident of its kind.

Have I Been Pwned confirms scale for first time

A new survey from the ed tech provider also found that two-thirds of instructors and students are concerned about artificial intelligence’s accuracy.

Teachers must lead the AI in education debate—setting classroom standards instead of letting armchair critics dictate them.

Anthropic launched Claude for Teachers last week: free for verified K-12 educators, standards-mapped curriculum, a K-12 Data Processing Addendum co-developed with the AFT. I'll say it plainly — the tool is effective, and this is the most privacy-forward launch of the classroom AI race so far.

OpenAI, Anthropic, Google, and Microsoft are investing in educators with free premium AI tools. Learn why teachers have become central to the future of AI in education

Calling Balls and Strikes on "Claude for Teachers" (because, well, that's what you expect from me, I know) I want to be clear about something I think matters: Anthropic intentionally built a purpose-built tool for educators. The Learning Commons connector, mapping to 50-state standards, integration of tools like MagicSchool AI, and the deliberate pedagogical scaffolding are impressive. That focus should be applauded, but it cannot come at the cost of student privacy. I am a fan of Claude (and use it a ton), but I am very disappointed in Anthropic’s entry into K-12 this week. Purpose-built matters, but so does understanding safety, security, and interoperability. Claude for Teachers launch feels like a major red flag regarding data privacy and infrastructure security. Here is where the fine print breaks down: 1. Bottom-Up Data Exposure: Anthropic is encouraging individual teachers to upload student rosters, diagnostic data, and attendance logs to analyze trends. These are enterprise data decisions that must be made by district leadership, not classroom teachers or a tech vendor bypassing IT governance. This is potentially and is a cybersecurity and data infrastructure nightmare. We've spent decades working to use standards to secure data and security - this is a miss. 2. Bypassing Secure Provisioning: To get verified for the free tier, teachers are asked to upload sensitive personal documents like IDs, paystubs, or employment contracts. Districts already have secure, established Single Sign-On (SSO) and rostering mechanisms to provision staff accounts. Why are we asking teachers to hand over sensitive personal data to a third-party processor just to access a tool? Adults need to protect their data too. 3. Zero District Visibility: Anthropic does not have a formal district-level dashboard or administrative offering ready at launch. This leaves school districts completely blind to what student data is being uploaded, how accounts are being managed, or where liability sits. Anthropic notes that their terms of service and K-12 Data Processing Addendum are “built around FERPA.” Let’s be precise: FERPA governs school districts, not technology companies. Claiming you are FERPA-compliant without giving districts the administrative architecture to actually manage and maintain that compliance is at the worst deeply misleading and at the best not aligned with best practices. We need to stop normalizing the bottom-up adoption of high-risk tools in schools. OpenAI took this path a year ago, but at the least, it provided a mechanism for districts to claim their domain and exercise institutional control (as you all know, I had lots to say there too). Google for Education has understood this too. My colleagues Mark Racine, Justin Reich, and others who have also highlighted some other challenges. Time to read the fine print, folks. We can protect teacher time without sacrificing student data autonomy. Graphic generated by Gemini. | 26 comments on LinkedIn

How one journalist found himself targeted by generative AI over a keyfob photo.

Lawyers must always keep their defense hats on to help identify the biases Generative AI may support and to implement creative safeguards to ensure the presumption of innocence is maintained in courtrooms.;