AI in Schools
Microsoft launches AI literacy credential ahead of Sydney conference
The Monday read is one credential launch that just changed the question schools should be asking about AI PD, one Sydney conference on Tuesday worth setting calendar alerts for, one privacy code clock with eighteen days left, and one quiet milestone in how Australian universities have voted with their settings.
Microsoft drops its AI literacy credential for educators this month
Microsoft Elevate for Educators launches its globally recognised AI Literacy for Educators credential in May 2026, co-developed with ISTE and ASCD and aligned to the European Commission, OECD, and Code.org AI literacy framework. The credential sits inside the broader Elevate program, the same program Microsoft is using to anchor the three million Australians it has committed to train by the end of 2028. For Australian schools it is the first ed-tech vendor PD that comes back to the school with an external badge against an external framework, not an internal completion certificate.
Why this matters: AI PD in Australian schools has run on goodwill, twilight sessions, and the occasional vendor masterclass. A credential changes the conversation in three places. First, principals can now write "credentialled in AI literacy" into a job advert and benchmark applicants against it. Second, performance and development plans can map a year of teacher work to something a registrar will recognise. Third, procurement teams comparing PD-bundled product offers from Microsoft, Google, OpenAI, and Adobe finally have a like-for-like axis. The risk is that the credential becomes another box-tick that crowds out school-context judgement, which is what AISNSW and ACER are quietly worried about. The work this fortnight is deciding whether the school will recognise the credential as part of its own PD framework, or treat it as additional. Microsoft Elevate for Educators, and Microsoft on the Australian three-million skilling commitment.
Sydney's AI in Education conference Tuesday, and what to listen for
Informa's Artificial Intelligence in Education NSW conference runs all day Tuesday 19 May at Swissôtel Sydney, with NSW Department of Education's Jason Miezis on generative AI implementation, ACARA's Melanie Hughes on AI literacy through the curriculum, and Macquarie's Professor Matt Bower on anticipatory governance for the AGI question. The agenda runs from 08:50 through 15:40 with assessment redesign, classroom integration, and ethical frameworks broken out as separate streams.
Why this matters: The interesting question on the Tuesday floor is not whether NSW is ahead (it is, on EduChat and on policy) but whether the curriculum side has caught up with the deployment side. Hughes's session is the one to attend if you are inside a state system, because the gap between Microsoft's classroom credential and what ACARA has formally embedded in the Australian Curriculum is the gap principals are going to be asked to close at the school level. Miezis's session is the one to attend if you are in an independent or Catholic system, because the NSW GenAI Safety, Capability, and Enablement framework is the closest thing to a national procurement pattern Australian schools have. For curriculum leads who cannot get to Sydney, the materials usually come back into the public domain inside two weeks. Informa AI in Education conference series.
The Children's Online Privacy Code clock is at eighteen days
The OAIC draft Children's Online Privacy Code closes for industry, civil society, and academic submissions at close of business Friday 5 June 2026, eighteen days from this morning. The code captures not just social media but online games, messaging services, educational tools, and streaming platforms that are likely to be accessed by children, which is to say most of the ed-tech estate a school deploys. The OAIC has been clear that the final code must be registered by 10 December 2026, which gives the regulator six months between submission close and registration to absorb the response.
Why this matters: The Canvas breach has done the school-system half of the work, because every CIO who walked the executive through that incident has the consent, data minimisation, and deletion conversation already open. The other half is the vendor side, and the question to put to every shortlisted ed-tech supplier this term is whether they have lodged a submission and what it says. A vendor that has not engaged in the consultation is telling you it has not modelled the operational change yet, which is a procurement signal in its own right. For schools, the practical work this fortnight is drafting a single-page school position to send to the OAIC, even a short one, so the regulator sees the school sector and not just the legal sector in the submission pile. Addisons on the consultation closing 5 June 2026.
Twelve Australian universities have now switched off AI detection
Curtin disabled Turnitin's AI writing-detection feature across all campuses on 1 January 2026, and a March stocktake puts the count at twelve Australian universities that have used Turnitin AI detection and then turned it off, including ANU, UQ, Macquarie, CSU, ACU, Swinburne, Tasmania, La Trobe, Griffith, Victoria, and UniSA. Fourteen universities continue to use it with documented caveats, including RMIT, which retains the detector as supplementary and pairs it with its in-house Val chatbot. The Curtin Academic Board's published reasoning, false-positive bias against international and EAL writers and a preference for assessment redesign over surveillance, is the document the other twenty-three universities are now reading against their own settings.
Why this matters: The detector-off cohort has crossed half the Australian university sector and the trend is now harder to reverse than to continue. For institutions that still have AI detection switched on, the question is no longer technical (does it work) but procedural (is your academic-misconduct process resilient if it does not). The harder operational work, redesigning assessment around demonstrable process rather than artefact, is what TEQSA's 2026 regulatory-led framework will start asking about by exception. Programs with high-stakes terminal essays are most exposed and should be the focus of the assessment-redesign budget for semester two. Curtin's update on the Turnitin AI detection tool, and the March 2026 stocktake of Australian university detector settings.
If you are at Swissôtel Tuesday, the cross-sector conversation worth having is what a credentialled, governed, detector-off Australian education system actually looks like in 2027. More on the operational thread at digitalattitudes.com.au.