A growing chorus of academics, union leaders and market analysts is urging a recalibration of the artificial intelligence conversation, arguing that the technology’s real dangers lie not in its capabilities but in how it is being deployed and understood. In Italy, more than 130 researchers from fields as varied as computer science, neuroscience and philosophy have signed an open letter calling for a return to “clear and realistic foundations” in public debate, pushing back against narratives of machine consciousness and wholesale human replacement. Viewed from Rome, the priority is not to halt innovation but to equip people with the literacy necessary to assess AI’s genuine limits.

This concern resonates in very different contexts. In Moscow, retail investors are flocking to AI-driven trading tools, with a recent survey by the broker eToro indicating that roughly thirty percent of individual investors now use AI for portfolio management. Some are training autonomous agents to trade on their behalf, occasionally boasting of astronomical returns that later prove fraudulent. Analysts in London note that such models tend to amplify the user’s own cognitive biases, often with disastrous consequences for risk management. The dream of a cash machine run by algorithms remains, for now, a dangerous illusion.

Meanwhile, from São Paulo, the anxiety takes a different shape. Journalism unions in the United States, Greece and the Philippines report mounting pressure as media companies integrate AI into editorial workflows—conducting research, producing transcripts, generating illustrations, and even writing articles. Yet according to a global survey of media managers, actual layoffs linked directly to the technology remain limited. More insidious, perhaps, is the cognitive toll on individual users. A spate of recent studies suggests that heavy reliance on generative tools like ChatGPT may erode creativity, attention span, critical thinking and memory. The very neural pathways that underpin independent thought, researchers warn, risk atrophying under the convenience of automation.

At the corporate board level, the promise of AI as a neutral advisor is equally fraught. In Russian boardrooms, directors often find themselves making decisions on matters outside their expertise, with insufficient time to consult external experts. AI could in theory provide a synthetic overview, but its outputs are only as reliable as the data and assumptions fed into them—and those are themselves shaped by the competing interests of shareholders, creditors and management that already crowd the table.

The common thread across these geographies is the human tendency to project onto AI powers it does not possess. As the Italian academics insist, the real imperative is education—teaching people to understand what the technology can and cannot do, rather than oscillating between hype and fear. Forward-looking policy, analysts in Brussels suggest, should focus on transparency, accountability and the preservation of human judgment. The machine may assist, but it cannot replace the messy, deliberate work of thinking.