Data released by the Politecnico di Milano Observatory outlines an unmistakable picture: the artificial intelligence market in Italy is accelerating sharply. With a forecast of 1.8 billion euros for 2025 and a 50% increase compared to the previous year, the experimental phase is definitively giving way to operational integration in corporate processes. The ICT sector is leading the way, aiming for an estimated productivity increase of around 25%.

The lowering of technical barriers through methodologies like "vibecoding" is triggering a frantic race for skills. Platforms like LinkedIn are recording a 112% jump in job postings for "prompt engineers", while roles like data analysts and AI engineers dominate recruiters' searches.

The flip side of the coin on the employment front manifests itself with cases like InvestCloud Italy in Veneto. The dismissal of 37 employees to favor an organizational model based on AI automation creates a heavy legal precedent, confirmed by the Court of Rome. This scenario requires a pragmatic reading: automating operational processes inevitably leads to reorganizing the workforce from the ground up. Learning to orchestrate complex workflows and create hybrid solutions becomes the only way to make oneself irreplaceable, fully confirming the triumph of orchestration over old manual tasks.

The evolution of workspaces and operating systems

Corporate data management is shaken up by the new developer platform launched by Notion. The famous workspace is transforming into a true hub for AI agents, allowing you to connect autonomous models, external sources, and custom code directly within documents and databases.

This architecture pushes the product well beyond the concept of a smart notepad. Making scripts and models work on operational data in the exact place where the team collaborates eliminates the need for cumbersome integrations or complex middleware. An agent can now scan support tickets and update product specifications in total autonomy, making the workspace a native execution environment.

On the mobile front, the integration of Gemini Intelligence announced during the Android Show 2026 brings proactive AI features into the core of the operating system. The device manages multi-step tasks like booking trips or dynamically filling out forms by acting in the background through apps like Chrome and Gboard.

The interaction paradigm evolves towards autonomy: the operating system interprets goals and completes them by coordinating installed software.

Native implementation at the operating system level sweeps away years of conventions tied to traditional graphical interfaces, decreeing the end of the wrapper era in favor of agents capable of orchestrating applications. Designing software today means facilitating the reading and manipulation of interfaces by artificial intelligence, ensuring interoperability and security.

Asynchronous development and the vendor cage

The software lifecycle accelerates further with the arrival of Codex on the ChatGPT mobile app. OpenAI paves the way for mobile development, allowing you to monitor pipelines, direct tasks, and approve the work of agents directly from your smartphone.

The idea of writing code from a phone remains impractical, but the engineering process is shifting towards supervision. Starting a complex refactoring, examining generated logs, and giving the green light for implementation with a tap transforms the developer into a reviewer. Evaluation metrics will increasingly shift to the ability to manage asynchronous flows and delegate complex tasks to machines.

In parallel, closed ecosystem dynamics are making themselves felt at Microsoft. Thousands of developers have had their licenses to use Claude Code, Anthropic's tool, revoked to favor the forced adoption of GitHub Copilot CLI. The directive imposes a return to proprietary tools for those operating in cloud environments managed by Redmond.

This sudden closure demonstrates the desire to lock down the programming interface market. For those managing technical teams, it means having to rework established automation pipelines. Evaluating open source alternatives or leveraging direct API calls becomes a strategic priority to avoid depending exclusively on the choices of a single vendor.

Financial research and automated aggregation

The European expansion of the new Google Finance marks a crucial step for market analysis. The platform abandons the simple aggregator format to integrate advanced research functions capable of answering complex questions about macroeconomics and specific stocks.

The "Deep Search" module offers conversational answers and in-depth links, working in synergy with interactive charts. Tapping a historical peak on a candlestick chart instantly generates a summary of the triggering event, bypassing hours of manual research on financial news outlets. Earnings conference calls become interactive thanks to live audio paired with synchronized transcripts and AI-generated summaries. This eliminates the need to maintain custom scrapers or Python scripts to extract textual insights, bringing advanced financial analysis directly into the browser.

In the tech information landscape, the return of Digg represents an interesting experiment in automated curation. Having abandoned community mechanics, the site focuses exclusively on tracking artificial intelligence news.

The engine under the hood ingests real-time content from X, applying sentiment analysis and signal detection techniques to identify relevant trends. If prominent industry figures interact with a post, the system tracks its impact and propagation. Delegating ranking to the engagement graph solves the problem of internal bots, but exposes the platform to a very high infrastructural risk: relying exclusively on the APIs of a single social network makes the business model extremely fragile.

Weekly radar: news and operational tools

To optimize daily workflows, in addition to the AI tools available on the site, the market continues to churn out solutions geared towards efficiency and automation. Here are the most relevant releases and corporate movements of the last few days:

• Grok Build: x.AI releases a programming agent executable directly from the terminal, designed to speed up code writing and refactoring in direct competition with OpenAI and Anthropic tools.
• Osaurus: a new native macOS application that allows you to run local AI models and seamlessly integrate them with cloud providers, maintaining the privacy of sensitive data.
• Vapi: the startup specializing in AI voice agents for customer support reaches a 500 million valuation, demonstrating the solidity of the voice-first market for enterprises.
• OpenClaw: a highly promising open source framework for building engineering workflows based on autonomous agents, useful for those who want to break away from closed-source logic.
• Cerebras: the hardware company marks the first major tech IPO of the year by raising 5.5 billion dollars, confirming the hunger for alternative infrastructures to traditional chips.
• Claude For Legal: Anthropic launches a version of its model optimized to automate document analysis and the work of lawyers.
• Proactive security: artificial intelligence autonomously discovers and blocks its first zero-day exploit on behalf of Google, marking a historic milestone for automated cybersecurity.
• Record valuations: Anthropic hits a 900 billion dollar valuation, temporarily surpassing OpenAI in the race for capital to train frontier models.