When Italy’s Senate approved its landmark AI law on September 17, it didn’t just make headlines - it drew a hard boundary. From now on, employers using AI systems in hiring, performance evaluation, or workplace monitoring must ensure transparency, human oversight, and explainability in all automated decisions (DLA Piper, DLA Piper).
Ireland’s Data Protection Commission released updated guidance that gently reminded everyone: GDPR still applies, even when the decisions come wrapped in shiny algorithms. Employers using AI in HR must continue to follow rules on fairness, data minimization, and transparency - even when it feels like “the machine knows best” (dataprotection.ie).
Meanwhile, on the other side of the world, New South Wales is working on proposed legislation that introduces a “digital duty of care” under its Work Health & Safety Act. If passed, it would restrict employers from using AI in ways that impose excessive workloads or unreasonable productivity metrics - and it would give unions or authorized personnel the right to inspect digital systems for compliance (DLA Piper).
In short: the global trend is clear. Regulators are no longer asking whether AI belongs in the workplace. They’re asking how, why, and who’s watching.
🚨 From hype to headache: why AI in HR is under scrutiny
It’s tempting to view AI as the great equalizer of modern HR - unbiased, efficient, always-on. But the reality is murkier. Many systems used for screening candidates, rating employee performance, or “boosting productivity” have been black boxes, offering little in terms of accountability or auditability.
The result? Rising distrust. Employees feel monitored but not supported. Decisions feel opaque. And teams start self-censoring - or worse, disengaging.
And this isn’t just theory.
A recent IAPP article highlighted that under the new EU AI Act, systems used in HR contexts are officially categorized as “high-risk.” That means organisations will be legally required to keep detailed documentation, run impact assessments, and provide recourse for human intervention.
In other words: HR tech is no longer “just a tool.” It’s infrastructure. And governments are watching.
🧠 Enter the Human-AI hybrid model
Let’s be clear: none of these laws say “don’t use AI.” They say use it responsibly. With documentation. With context. With the people it affects actually involved in the process.
Which, to be honest, we’ve always believed.
At Ulla Technology, we design our onboarding and AI deployment processes with human oversight, clarity, and compliance in mind - not as an afterthought, but as a principle.
💬 Before an HR team starts using AI-powered meeting analytics with Ulla HR, we walk them through what’s measured, why it matters, and how it benefits their people. It’s not about “monitoring.” It’s about getting visibility into communication patterns, meeting clarity, and signs of fatigue - without watching every move or tracking keystrokes.
🛠️ In Italy, the new law even requires the appointment of a responsible person to oversee AI systems in workplace contexts. We support that fully. AI should support HR teams and leaders - not replace them, not override them.
Because if your AI system makes decisions faster, but your employees trust you less, you didn’t automate progress. You just outsourced accountability.
💡 What responsible insight actually looks like
Tools like Ulla HR are built to offer meaningful insight into how teams work - not to control them. And yes, you can use Ulla Notetaker or any other meeting capture tool as input - we process transcripts, voice patterns, and metadata to offer:
- Team-level communication clarity: Are action items clear? Are ideas left unresolved?
- Speaking balance & attention drift: Does one voice dominate? Are meetings collaborative or lopsided?
- Signs of disengagement: Not from what someone types, but from how they participate.
- Burnout signals: Subtle, measurable trends - like rising passivity or dropping meeting sentiment.
These insights never replace humans. They inform them - so leaders can step in with empathy, not guesswork.
🤖 So... What happens when AI becomes the Minister?
Fun fact: In September, Albania appointed the world’s first AI Minister - an actual virtual persona named “Diella,” tasked with public communication, policy updates, and helping coordinate digital reforms (Politico).
Sure, it’s partly a PR move - but it also reflects the deeper reality: AI is no longer hiding in the basement, running models. It’s showing up in the meeting room, the HR dashboard, the leadership report. It’s wearing a blazer now.
And while we at Ulla may not be gunning for a ministerial position (yet 😉), we believe AI belongs in workspaces as a co-pilot - one that informs, clarifies, and saves time, not one that quietly replaces trust with surveillance.
🪄 Final word
The future of work isn’t AI instead of people. It’s AI with people - smarter tools for better conversations, healthier teams, and faster, fairer decisions.
And if you’re exploring how to bring that future into your team (without ending up on a regulatory watchlist), we’re here for it.