This week we’re running a small editorial experiment. Do you enjoy reading Above Trends, Beyond Insights at the end of a busy Friday or during a slower Saturday morning? We are curious to know what feels most natural for our readers, so if you are reading this newsletter over the weekend, don't be too surprised.

Europe’s shifting AI Act: from January certainty to a new phase of regulatory recalibration

In the beginning of last year, the European Commission released the first implementation guidelines for the AI Act. They outlined how risk classification should work, what documentation high risk providers would be expected to maintain and how public sector deployments should be governed. For many organisations this became the basis for early compliance planning.

But over the past days a different picture has emerged. Several reports highlight a growing political appetite to revisit not only the guidelines but also some interactions between the AI Act and the broader Digital Omnibus regulation. The language used by regulators has become more cautious and some member states have signalled concerns about administrative burden, innovation slowdowns and the compatibility of new obligations with existing national frameworks.

Although there is no formal reversal of the AI Act itself, the tone suggests that enforcement phases, secondary legislation and guidance documents may undergo revision or delay.

Why now? One hypothesis is that the rapid acceleration of foundation models and sector specific AI has made the previous guidelines appear too narrow or overly focused on pre generative AI categories. Another factor is geopolitical competition. Several governments fear an innovation drain if compliance is perceived as excessively heavy in comparison with the United States or East Asia.

Budgetary pressure also plays a role. Implementing supervisory authorities capable of auditing high risk systems is expensive and politically sensitive. A final element is electoral. As Europe approaches a new political cycle, regulatory positions are shifting to accommodate national priorities and coalition negotiations.

The practical implication is to resist waiting for perfect clarity. Investing in internal governance, data quality, traceability and model documentation will remain essential regardless of regulatory tempo. Preparing now creates optionality. Waiting increases risk.

TREND TRACKER

Carbon aware cloud scheduling

Carbon aware cloud scheduling is moving from experiment to practice in large organisations that run significant workloads in the cloud. The core idea is simple. Instead of executing compute jobs whenever capacity is available, systems use real time or forecast data about grid carbon intensity to decide when and where to run non urgent tasks. Google has publicly described how its “carbon intelligent computing platform” shifts data centre workloads to hours and locations with more wind and solar on the grid, without adding new hardware or degrading service quality. The Green Software Foundation has codified this logic into a Software Carbon Intensity standard and a Carbon Aware SDK that developers can use to move or delay workloads based on carbon signals. Academic and industry research shows that such scheduling can significantly reduce emissions for energy intensive services by adapting workload quality and timing to grid conditions. At the same time, independent guides and tooling ecosystems are emerging to help enterprises integrate carbon aware decisions into pipelines, schedulers and orchestration platforms. For European leaders under growing sustainability reporting pressure, this trend matters because it turns “green IT” from a procurement question into an operational capability: a way to reduce emissions, demonstrate concrete progress and align technology strategy with regulatory and stakeholder expectations.

Want to learn more?

• Climatiq - A guide to carbon-aware computing

• GSF - How UBS succeeded in measuring and reducing carbon emissions of their core risk platform

• Google blog - We now do more computing where there’s cleaner energy

QUICK INSIGHT

Three patterns to cut cloud costs without slowing innovation

Cloud budgets tend to grow faster than the systems they support. Many teams discover cost inefficiencies only when finance raises a flag, yet most savings come from operational hygiene rather than structural redesign.

• Implement automated kill switches for non production environments and idle workloads. This alone can cut monthly waste by double digit percentages.

• Monitor anomalous consumption with threshold based alerts that surface sudden scaling, misconfigurations or runaway processes.

• Use time based job scheduling to execute batch tasks during low cost periods or in lower carbon intensity windows if available. This improves efficiency and sustainability at the same time.

WHO IS BABINI MAZZARI

Our Value Proposition

Babini Mazzari is the strategic IT partner for European companies looking to navigate digital transformation in a structured, pragmatic, and sustainable way.
We don’t just deliver technical solutions - we work as an extension of your internal team, helping you integrate systems, optimize processes, and lead change with clarity and competence.

Our approach is built on listening, transparency, and a strong results-driven culture. Whether you're scaling, modernizing, or rethinking your operating model, we support every client with the right tools, clear methodology, and long-term vision.
Above Technology. Beyond Solutions.

📎 [Our services]
📎 [Our Manifesto]