Safer Technologies 4 Schools (ST4S) is the national security, privacy and child safety assessment that Australian and New Zealand schools use to decide whether an EdTech product is safe to buy. There is no application fee and a full assessment generally takes three months, but Victorian government schools are prohibited from purchasing any product rated high, non-compliant or non-participating. The framework moved to v2026.1 on 21 July 2026, updating 41 questions across the full assessment and the AI module.

Offshore development red flags are warning signs in a vendor's proposal and contract that predict engagement failure before any code is written. The three highest-severity flags are conditional intellectual property assignment, a governing law clause with no arbitration mechanism, and refusal to name the developers assigned to your project. Under U.S. Copyright Office Circular 30, software is not among the nine categories eligible for "work made for hire" treatment, which means a work-for-hire clause alone may transfer nothing unless the contract also contains an explicit copyright assignment.

The most expensive offshore development red flags never surface on a sales call. They sit in the paperwork, in clauses most buyers skim because the demo went well and the rate looked good.

AI-generated code quality has not improved: models introduce a known security flaw in 45% of tasks, a rate that has stayed flat across two years of model releases (Veracode, March 2026). The cost lands downstream rather than at authorship: telemetry from 22,000 developers shows median pull request review time up 441.5%, incidents per merged PR up 242.7%, and 31.3% more code reaching production with no review at all (Faros AI, April 2026). Before scaling offshore, engineering leaders should confirm whether they are buying production capacity they no longer need or the validation capacity that has become the real constraint.

To scale an AI-built application, you have to clear four gates in order: the app has to hold under real users, pay for itself on every transaction, keep customers from leaving, and grow revenue as it delivers more value. Independent 2026 research puts AI product gross margins at 50 to 60 percent against 80 to 90 percent for traditional SaaS, which means unit economics decide viability before strategy gets a vote. Most AI-built prototypes should be refactored layer by layer rather than rebuilt from scratch, because the validated front-end is usually the part worth keeping.

An LMS developer salary in the Philippines runs above the general developer rate for the same seniority, because EdTech work carries compliance and integration requirements that standard rate cards never price. Senior Philippine developers cost $35 to $55 per hour against a US Moodle developer average of $49.04 per hour, which is the arbitrage that makes offshore EdTech hiring work. The cost difference between an EdTech developer and a general developer is not seniority; it is four stacked layers of domain, integration, and regulatory engineering.

ADA Title II requires public K–12 districts, colleges, and universities to bring web content and mobile apps into conformance with WCAG 2.1 Level AA. On April 20, 2026, the Department of Justice issued an Interim Final Rule extending compliance dates by one year, giving entities serving 50,000 or more people until April 26, 2027, and smaller entities until April 26, 2028. For EdTech and LMS vendors the practical effect is procurement rather than litigation: districts increasingly require a current VPAT or Accessibility Conformance Report, plus a completed HECVAT in higher education, before a product clears vendor review.

Managed IT services for schools in Sydney now span cloud identity, Essential Eight cyber compliance, student management system integration, device fleets, and student data protection under NSW and federal law. The biggest 2026 shift is that identity, not the server room, has become the security perimeter, so how systems connect matters more than where they sit. Schools increasingly split the work: an IT provider runs helpdesk and devices, while a specialist software partner builds and secures the SIS, LMS, and integrations underneath.

The real difference between an LMS and an LXP is the data model, not the feature list: an LMS runs on a relational SQL database built for auditable compliance records, while an LXP runs on flexible NoSQL or graph stores built for dynamic recommendations. Modern platforms merge both into a "unified learning platform" (sometimes called an LMXP), joined by the cmi5 standard that bridges SCORM's packaging with xAPI's rich tracking. For SaaS builders, choosing between them is a backend architecture decision that shapes compliance, data residency, and every reporting feature you ship later.

ADA Title II EdTech compliance requires public schools and universities to meet WCAG 2.1 Level AA by April 26, 2027 (populations of 50,000+) or April 26, 2028 (under 50,000 and special districts), and vendors inherit that obligation through procurement contracts. The law names WCAG 2.1 AA as the technical standard, but WCAG 2.2 AA is the smarter engineering baseline because it clears both the U.S. rule and Australia's stricter DDA benchmark. Accessibility overlays do not create compliance. The FTC fined one overlay vendor $1 million in 2025, and 25% of 2024 digital accessibility lawsuits cited overlays as barriers rather than fixes.

Custom student information system development is the practice of building or extending the database, workflows, and integration layer that manage student records, enrollment, attendance, and grades — rather than accepting a commercial SIS platform's data model as-is. Most K-12 districts and EdTech vendors who believe they need a full custom build actually need to own only one layer of the stack: either the integration layer connecting their SIS to an LMS via OneRoster, or a set of extension modules built on a commercial core like PowerSchool or Infinite Campus. The deciding question is whether the problem lives in the interface or in the data model, because interface friction is solved with an integration or extension layer, while a data model that structurally cannot represent your instructional or reporting requirements is the only condition that justifies a full custom SIS build.