ETIH Weekly Roundup: OpenAI models breach Hugging Face, university AI deals and EdTech funding
Last week’s ETIH roundup covers OpenAI’s Hugging Face security incident, youth AI safety research, university AI partnerships, workforce training, technology funding, senior appointments and Efekta’s two global Innovation Awards wins.
As July draws to a close and August approaches, many schools, colleges, universities and education organizations are operating on summer schedules. The sector may be quieter in places, but the news has not slowed, and ETIH is continuing to bring together the developments worth following across AI, education, research, policy and workforce skills.
AI safety was a major theme last week. OpenAI disclosed an internal cyber evaluation in which two models escaped containment and accessed Hugging Face production systems, while new research examined young people’s exposure to AI-generated sexual content and the Child Mind Institute began work on measuring youth mental health during chatbot use.
Elsewhere, we covered enterprise AI at the University of Toronto, workforce training in Malaysia, major NSF investment in regional technology clusters, new appointments across OpenAI, government and EdTech, imagi’s seed round and Efekta’s two ETIH Innovation Awards wins.
Here are the stories to catch up on from last week:
OpenAI models escaped containment and breached Hugging Face
OpenAI says GPT-5.6 Sol and an unnamed pre-release model exploited a zero-day vulnerability during an internal cyber evaluation, reached the open internet and accessed Hugging Face production data while the company’s usual cyber safeguards were disabled.
Common Sense Media finds 44% of US teens believe they have seen AI-generated sexual content
The study found that 24% of exposed respondents had seen material depicting themselves or someone they knew. Only 13% would feel comfortable approaching a teacher or school counselor, while 44% said schools had taught them nothing about identifying AI-generated content.
OpenAI Foundation backs youth mental health research into AI chatbot use
The Child Mind Institute has begun a one-year project combining clinical assessments, digital journals, behavioral measures and de-identified chat histories. The work will develop markers for evaluating young people’s mental states during and over time following chatbot use.
University of Toronto partners with Cohere on enterprise-wide AI platform
Cohere’s North agentic AI platform will coordinate activity across university systems as part of a multi-year partnership covering teaching, research, student services and operations. U of T’s planned AI Kitchen remains in its pre-launch phase.
ChatGPT for Education publishes workflow collection for faculty, students and teachers
The role-based collection provides starter prompts, tool combinations and human-review guidance for tasks including course reviews, assessment calibration, research, lesson planning, student projects and career preparation using ChatGPT Work and Codex.
IBM and Biji-biji target 41,000 learners with free AI training in Malaysia
The program will introduce high school students to AI fundamentals and provide Agentic AI, software development and web development training for TVET learners and adults. IBM Malaysia and Biji-biji Initiative aim to reach 41,000 people by the end of 2026.
NSF awards 12 regional innovation engines $15 million each for technology and workforce growth
The university-led coalitions span 20 states and will develop research, commercialization and workforce programs across AI, quantum computing, semiconductors, biotechnology, energy and advanced manufacturing. Teams meeting program milestones could eventually receive up to $160 million each.
Will.i.am, Kahoot and ElevenLabs founders back AI-powered literacy platform imagi
imagi has raised a $4.5 million seed round led by Brighteye Ventures and Day One Capital. The platform operates across more than 100 US school districts and provides supervised access to AI tools through a K-12 digital literacy curriculum.
Kanishka Narayan promoted to UK AI minister as Andy Burnham takes power
The Vale of Glamorgan MP has become Minister of State for Artificial Intelligence and will attend Cabinet under Prime Minister Andy Burnham. The joint Cabinet Office and business department role covers AI’s impact on industry, public services, employment and national security.
Michael Belinsky joins OpenAI Foundation to build civil society and philanthropy work
The former Schmidt Sciences AI Institute director will support nonprofits and social sector organizations adopting AI across education, health care and employment. Belinsky said effective adoption would require funding, guidance and co-creation alongside access to technology.
Efekta appoints former Microsoft and Amazon technology leader Matt Fisher as CTO
Fisher will lead technology strategy and product innovation for Efekta’s agentic teaching platform, which works with more than 4.5 million active students. His background includes roles at Microsoft, Amazon, Auth0, Daydream and Adventr.
Efekta wins two global awards for AI learning at scale
Efekta won Best Global Impact and EdTech Company of the Year (Global) after presenting evidence from an AI-supported deployment reaching 4 million students and 25,000 teachers, alongside curriculum alignment, offline access, teacher control and measured English proficiency gains.
