Why Italian universities need a verified skill graph — now

1. The EU AI Act high-risk regime lands on employment AI in December 2027

Regulation (EU) 2024/1689 — the AI Act — classifies AI systems used to evaluate candidates for employment as high-risk under Annex III, point 4. Those obligations apply from 2 December 2027: Regulation (EU) 2026/1744 deferred them from 2 August 2026. (Entry into force and application are different things — the AI Act itself has been in force since 1 August 2024.) Key duties:

• Transparency on the system's purpose, inputs, and limitations
• Human oversight in place, with authority to override
• Traceability / audit logs of decisions
• Right to explanation for affected individuals (Art. 86)
• Data governance (Art. 10) — training data lineage + bias testing

What this means for universities: the Act places distinct duties on the provider of a matching system and on the institution that deploys it, and the scope reaches third-party recruiter platforms a career service recommends. Which duties apply to you is a question for your DPO against the text of Regulation (EU) 2024/1689 — not one a vendor should answer for you.

2. What ANVUR cycles ask for: documented stage outcomes, traceable by programme

Italian universities face periodic ANVUR evaluations for accreditation. The quality of placement data, curriculum-to-labour-market alignment, and documented stage outcomes factor into the final scoring. The gap between what reviewers want and what most career services can produce is widening every year:

• Placement data freshness — end-of-year surveys with 15% response rates yield data 6–12 months stale by submission
• Skills alignment evidence — hard to produce without a verified skill graph and labour-market overlay
• Stage outcomes documentation — paper conventions and email threads are slow to reconstruct at review time

InTransparency's Skills Intelligence dashboard produces the exact evidence reviewers look for: program-level gap indices, alignment scores, 12-month trends, stage→hire conversion, time-to-placement. Export-ready.

3. Manual tracking carries a cost that is rarely measured

We modelled the cost of running an Italian career service without verified-credential infrastructure. For a typical 10,000-student university with 5 FTE career staff and 120 events/year, the hidden costs break down as:

• ~4,000 staff hours/year on manual placement tracking, stage paperwork, survey chasing, event logistics
• ~€116,000/year in fully-loaded labour costs reallocable to higher-value work
• €8,000+/year in subscription costs for placement tools and survey platforms

InTransparency is free for universities. The platform is funded by employers who pay for verified-evidence matching — universities reclaim budget and time in the same motion.

4. Reputation is the long-term dividend

Every verified student profile carries the issuing university's signature, and the student can export it as a signed credential in Europass/EDCI JSON-LD — the format Europass readers and EU Digital Wallet clients consume — verifiable by anyone against our public key. Your institution's name travels with the evidence, in a shape built to be read across the EU.

In parallel, international ranking bodies — QS Graduate Employability, THE impact scores, U-Multirank — are rapidly shifting weight toward verified placement outcomes. A university that can produce verified data today has a compounding advantage over one that starts from surveys.

Prospective students and parents read employability stats. A public university scorecard with verifiable projects, verifiable endorsements, and verifiable placements is the most credible recruitment asset you can publish.