The conversation about AI in VET has changed quickly. Six months ago, most discussions centred on which tools to adopt. Today, the focus is on governance, organisational capability and embedding AI in ways that improve outcomes without creating new risks. 

Days later, ASQA released its Principles for the Responsible Use of AI in VET, reinforcing many of the same questions the sector's leading institutions were already working through. 

The governance gap is real  

Our 2026 AI and Innovation Benchmark, which surveyed almost 100 education organisations across VET and higher education, shows AI has quickly become mainstream. Ninety-five per cent are already exploring or using AI, 28% have embedded it into at least one core workflow, and 71% expect to increase investment over the next 12 months. 

Governance, however, hasn't kept pace. Only 34% of organisations have a documented AI use policy and just 30% have a responsible AI framework, despite ongoing concerns around human oversight (55%) and sensitive institutional data being shared with external AI models (50%). That's exactly the gap ASQA's new principles are designed to address. 

What responsible AI looks like in practice 

Rather than introducing new regulatory obligations, ASQA's principles provide a practical framework for applying the 2025 Standards for RTOs as AI becomes part of everyday operations.  

The strongest message running through the guidance is human accountability. ASQA is explicit that AI should support professional judgement, not replace it. Trainers and assessors remain responsible for decisions affecting students, and providers need to demonstrate not only where AI is being used and who is accountable, but that staff can recognise when AI outputs are incomplete, inaccurate or simply wrong. 

One question is particularly relevant for technology leaders. Under Principle 3, ASQA asks whether your Student Management System or Learning Management System includes AI capabilities and, if so, how those capabilities are governed, where student data is processed and whether issues such as data sovereignty have been considered. Increasingly, those questions won't just shape internal governance - they'll influence how providers evaluate technology partners. 

Leading providers were already heading in this direction 

Listening to Cathy and David at Ready Summit, it became clear that many of the governance practices ASQA has now published were already taking shape across the sector. 

Cathy described Melbourne Polytechnic's decision to establish governance before scaling AI. As staff enthusiasm grew, the organisation responded by establishing an AI Steering Committee chaired by the CEO, surveying more than 2,300 students and 350 staff to understand AI readiness, and introducing a two-lane assessment model distinguishing where AI can support learning from where competency must still be demonstrated independently. 

David outlined a similar response from UNSW College. Staff were already experimenting with AI tools, so the focus became providing clear guardrails rather than restricting adoption. The team introduced an AI asset register, guidance on approved tools and data handling, and practical education to help staff understand where AI adds value, where human judgement remains essential and where the risks sit. 

Although they approached the challenge from different parts of their organisations, they arrived at the same conclusion. Governance isn't a brake on AI adoption - it's what allows institutions to scale AI with confidence. 

Orqestra: AI built for the realities of VET 

The same thinking has shaped how we've approached AI at ReadyTech. Every provider already holds enormous amounts of data across its student systems, learner records and compliance platforms. The challenge isn't collecting more information - it's making it accessible when people need it. For us, that's the shift from being data burdened to data powered. Orqestra was built to enable that shift by connecting intelligence across an institution's technology ecosystem, bringing information, automation and action together in a single experience. 

Whether preparing an AVETMISS submission, responding to a regulator audit or compiling executive reports, staff can ask questions in natural language and receive traced, auditable answers in minutes rather than days. One pilot customer reduced audit preparation from three to four weeks to just a few hours. 

Beyond compliance, education teams are using Orqestra for enrolment reporting, student data validation and recurring operational processes. Instead of cleaning up thousands of records before an audit, teams review a small number of exceptions each day. 

Just as importantly, Orqestra is designed to keep your people in the loop. It surfaces information, identifies patterns, recommends actions and automates routine tasks, but decisions remain with the people responsible for students, compliance and institutional outcomes. AI reduces the administrative effort around decision-making without replacing the professional judgement that sits at its centre. 

Governance by design 

Governance isn't something we've layered onto Orqestra after the fact. It's been built into the platform from the beginning, shaping both how AI capabilities are developed and how they're used in practice.  

New AI features are reviewed through ReadyTech's internal AI Council before release to assess risk, security and responsible use, while institutions retain control over model selection, permissions and data residency. Customer data is never used to train foundation models, and ReadyTech's AI governance is underpinned by ISO/IEC 42001 and ISO/IEC 27001 certification. A central command centre gives leaders visibility over AI usage, adoption, cost, risk and business impact, ensuring AI can scale across the organisation without compromising governance or accountability. 

As AI becomes embedded across VET, institutions will increasingly judge technology not just by what it can automate, but by how responsibly it has been designed, governed and deployed. The providers best placed to adopt AI at scale will be those whose technology combines productivity with accountability, human oversight and trust. 

Resources 

ASQA's full Principles for the Responsible Use of AI in VET, self-assurance questions and case studies are available on the ASQA website. 

You can also access the full findings from ReadyTech's 2026 AI and Innovation Benchmark. If you're reviewing your AI strategy, exploring the implications of ASQA's new principles or would like to see how Orqestra supports enterprise AI in regulated education, we'd welcome the opportunity to continue the conversation.