Frontier AI audits, quantum budgets and a bilingual classroom: five tech moves worth knowing this week
The week's announcements shared a common thread. Across identity security, cryptography, cloud infrastructure and the classroom, organisations are taking more direct ownership of the technology that protects them and the skills needed to understand it. Some of that is defensive, as AI makes software weaknesses cheaper to find and quantum computing threatens today's encryption. The rest is about capability: building, testing and teaching the systems locally rather than buying them sealed.
Abu Dhabi's Technology Innovation Institute (TII) appears in three of the five stories, as a research partner, a report sponsor and, through its chief executive, a leader of QuantumGate. That concentration is telling. Governments and enterprises worldwide are treating cryptography and AI security as strategic capabilities, and the UAE is one of the clearest examples of a country building a research base to match.
1. Delinea is pointing Anthropic's most capable model at its own credential vaults
Delinea, the identity security platform that governs what AI agents, humans and machines can do once they have access, has joined Project Glasswing, Anthropic's initiative to secure critical software using frontier AI. The company is testing Claude Mythos 5.1 for defensive cybersecurity work, starting with the code that stores, brokers and rotates privileged credentials for its customers.
"AI is making it much easier to find weaknesses in software, and that changes what customers should expect from security companies," said Art Gilliland, CEO at Delinea. "We joined Project Glasswing to use the best tools available to attack our own assumptions and find problems before someone else does."
The scope covers vaulting logic, secret storage and rotation, key management and cryptographic implementation, along with the session-brokering and proxy code that sits between a user or agent and a target system. Delinea is also examining authorisation enforcement, where a logic flaw in policy evaluation can be as damaging as a memory-safety bug. Findings will pass through the company's existing product security process, and recurring patterns will become secure coding standards, pre-merge checks and test coverage.
Pierre Mouallem, Chief Information Security Officer at Delinea, said the model is governed the way the company governs any other powerful account. "We're treating Mythos like any other privileged identity: its access is scoped, temporary and logged," he said. "It runs against our code in an isolated environment with no customer data in scope. Anything it finds enters the same product security and disclosure process as a human-reported vulnerability."
The significance reaches well beyond one vendor. Identity platforms hold the keys to every system they govern, so a single flaw in them is a flaw everywhere. If attackers can use AI to find weaknesses faster, security vendors that do not audit themselves with equivalent tools will struggle to justify their customers' trust. Delinea will share validated findings and remediation patterns with the Glasswing cohort and follow coordinated disclosure timelines for third-party and open-source components.
2. Nearly two-thirds of surveyed organisations have already put money behind post-quantum migration
Some 64% of organisations surveyed in key GCC markets have a budgeted post-quantum initiative, and 84% are planning a transition, according to The State of Post-Quantum Readiness in Key GCC Markets, published by QuantumGate and the UAE Cyber Security Council with support from the National Cryptography Center and TII.
Awareness is high. Some 97% of respondents recognise the harvest-now, decrypt-later threat, in which adversaries collect encrypted data today to break once a capable quantum computer exists, and 58% believe such a machine could emerge within five years. H.E. Dr. Mohamed Al Kuwaiti, Head of Cyber Security for the UAE Government and Chairman of the UAE Cyber Security Council, said: "As digital services, critical infrastructure and data become increasingly connected, quantum-safe cryptography must become part of the fabric of our national cybersecurity ecosystem. Our focus is on giving organisations a clear framework to assess their cryptographic environment, plan early and move forward in a coordinated way."
Execution lags intent. While 66% of respondents say they have visibility into the cryptography they use, only 35% have begun cryptographic discovery, the inventory of keys, certificates, algorithms and protocols that tells an organisation what needs replacing and in what order. Planning splits evenly, with 42% working to a defined roadmap and another 42% still building one, and only 34% rate themselves well prepared to carry out the transition. Supplier choices also remain open: 58% are evaluating several providers and just 6% have selected one.
Dr. Najwa Aaraj, Chief Executive Officer of QuantumGate and TII, said: "Funding is being committed and planning is already under way. The next step is to translate that intent into an executable programme. That starts with knowing where cryptography sits, what it protects and where the greatest exposure lies."
The finding with the widest implications is that 98% of respondents require cryptographic solutions that are nationally governed, locally controlled, or both. The report concludes that sovereign control has become a baseline procurement requirement in these markets. For global vendors selling post-quantum products since NIST published its first standards in 2024, that shifts the competitive question from algorithm choice to who holds the keys and where.
3. Core42 and TII want sovereign cloud secured with cryptography built in the UAE
Core42, the G42 company specialising in AI infrastructure and sovereign-enabled cloud, signed a strategic collaboration agreement with TII at GISEC Global 2026 to co-develop, evaluate and deploy security technologies across its sovereign-enabled cloud and Compass, its generative and agentic AI platform. The scope includes hardware security modules, confidential computing, encrypted AI inferencing, secure key management, encryption and trusted computing.
Rajeev Nair, Senior Vice President, Special Projects at Core42, said: "Sovereignty means more than deciding where data and infrastructure reside; it is also about building the technology capabilities required to protect them from all angles. Working with TII gives us an opportunity to evaluate and advance UAE-developed sovereignty and cybersecurity technologies against the requirements of sovereign-enabled cloud and AI environments."
Read alongside the QuantumGate research, the deal shows supply and demand moving together. If nearly every organisation surveyed wants locally governed cryptography, someone has to build and prove it on production platforms. Dr. Victor Juan Mateu, Chief Researcher at TII's Cryptography Research Center, said: "Our collaboration with Core42 creates an important pathway to evaluate and advance locally developed cybersecurity and trusted computing technologies within sovereign cloud and AI environments. By bringing together deep research expertise and real-world platforms, we can accelerate the journey from breakthrough innovation to practical deployment."
4. Presight is opening a route from South Korean AI companies to markets across three regions
Presight, the ADX-listed applied AI company majority-owned by G42, signed separate agreements in Seoul with the Gyeonggido Business & Science Accelerator (GBSA), a public institution supporting SMEs, and Dtonic Corporation, a South Korean AI operating system developer. Signed during the Abu Dhabi Investment Forum, the deals cover start-up deal flow, acceleration and cross-border market access, with Presight introducing selected Korean companies to partners across the Middle East, Africa and Central Asia.
Hyun-Gon Kim, President of GBSA, said: "Through this collaboration with Presight, we aim to create stronger connections between our respective technology ecosystems and provide promising AI companies from South Korea with access to new partners, investors and international market opportunities."
Dtonic, a member of Cohort II of Presight's Accelerator Programme, builds an AI OS around ontology, autonomous AI and explainable AI, with uses across smart cities, industry, the public sector and defence. Its agreement opens the possibility of that technology being used in Presight's current and future projects. Yong Joo Jun, Chief Executive Officer of Dtonic, said: "This agreement marks an important next step in our relationship with Presight. We look forward to exploring how Dtonic's AI OS capabilities can complement Presight's work in smart cities and data infrastructure."
The move reflects a wider pattern in which Gulf AI companies act as distribution channels for technology developed elsewhere, pairing capital and government relationships with specialist engineering from markets such as Korea. Thomas Pramotedham, Chief Executive Officer of Presight, said: "Our focus is on how these capabilities can help clients make better decisions and translate those decisions into action. By exploring practical applications together, we aim to build partnerships that support trusted intelligent systems and create lasting value for the organisations and communities they serve." Presight will host its inaugural Future Forum in Abu Dhabi on 5 October, focused on how AI that can decide and act can be trusted and governed.
5. iCodejr's Bloqi bets that schools want one coding and AI system from age four to 18
UAE-based digital and AI literacy academy iCodejr has launched Bloqi, a B2B platform that lets schools teach coding, AI and robotics through a single bilingual environment for the 2026-27 academic year. Schools buy one annual per-student-per-site licence covering early-years picture blocks, a visual Block IDE, an AI and machine learning sandbox where older students train and test their own models, and browser-based robotics. English and Arabic are built in throughout, and the system works offline.
Hannan Moti, Co-Founder and CEO of iCodejr, said: "Schools know their students need to understand AI, but many do not have a computer science specialist or the confidence that a tool they buy today will still meet their needs in three years. Bloqi gives them one system that can grow with the student and is designed so that a non-specialist teacher can begin delivering lessons quickly."
The launch follows the UAE's introduction of AI as a formal subject from kindergarten to Grade 12 in 2025-26, and sits within a global move toward deeper AI literacy. UNESCO's AI Competency Framework for Students describes young people as co-creators of AI, and the World Economic Forum's Future of Jobs Report 2025 names AI and big data, networks and cybersecurity, and technological literacy as the three fastest-growing skill areas through 2030. Moti said classroom AI has to go deeper than chatbot use. "Teaching a child to prompt a chatbot is not the same as teaching them AI," he said. "When they train a model themselves, they begin to understand the relationship between data, decisions and outcomes, including where bias and errors can enter the process."
The teacher shortage behind the product is a problem schools share worldwide, and Bloqi's bet on non-specialist delivery addresses it directly. "Every design decision traces back to something that was not working for a child or teacher in a real lesson," Moti said. "Schools should not need to assemble several products simply to offer a coherent computing programme."