We are unlocking intelligence for all but we have forgotten to unlock our own empathy. In March 2025, social media was aflame with the “Ghiblified” craze – millions of users transforming selfies into Studio Ghibli–style reveries, overwhelming OpenAI’s servers and prompting pleas from Sam Altman to “chill out”.
 
Yet, that same month, Meta sued Joy Timeline HK Limited, over ads for its CrushAI “nudify” app – an AI tool that digitally stripped strangers without consent, weaponizing tech used for curiosity into humiliation and blackmail
   
This stark juxtaposition between wonder and exploitation lays bare, the real fault line of our AI era. We have left behind, the age of calculators, tools that simply crunched numbers, into an era, where anyone can conjure a voice that mimics a loved one, craft bespoke lessons for students, model climate scenarios in hours rather than months, or automate entire workflows from drafting reports and legal briefs to summarizing research and streamlining supply chains, boosting productivity once thought years away and promising even greater gains.
  
Yet, that same technology that ignites breakthroughs, personalises education and supercharges efficiency can also be turned against us, fulfilling “remove her clothes” prompts to undress women without consent, or just last week, applying synthetic “glue” dripping down a user’s face in a non-consensual image creation. This fragile boundary between empowerment and exploitation demands we rethink, not just what our tools can do but who we become when we wield them.
 
From AI-safety teams drafting red-teaming frameworks to ethicists mapping out “provably benign” models, the community has produced a mix of proposals. Meanwhile, policymakers chant “Just regulate” as if that slogan could outpace every new model and funding round. In truth, legislatures move at a very slow pace compared to GitHub commits and capital flows. By the time a comprehensive bill reaches the floor, lobbyists have hollowed out its core; before regulators can even agree on a “fair use” framework for synthetic media, bad actors have decamped to unmonitored forums. Worse, over-rigid mandates risk freezing out the scrappy startups and pioneers that powered AI’s democratic ascent – handing the spoils to hyperscale firms with armies of compliance attorneys and cementing their moat, or state that could not be bothered to play by the same rules.
  
The crisis is threefold – legal, technological and cultural. Lawmakers struggle to define and police AI’s boundaries, leaving loopholes that bad actors exploit. Meanwhile, the technology itself evolves faster than our capacity to anticipate its misapplications, spawning tools that outpace safeguards. And at the heart of it all lies a cultural rot; we have become thirstily eager to humiliate one another for clicks, to prize virality over veracity, and to amplify disinformation simply because it “feels okay” in our feeds.
  
AI didn’t invent cruelty but it has handed us instruments to carve our worst impulses into the digital ether. Rather than slam the brakes on AI, whose promise in various fields is too great to abandon, we must ask what social architecture and laws will channel these capabilities towards empathy and cultivate digital citizenship before every screen hosts an autonomous agent.
  
Policymakers should shift from reactive bans to designing guardrails that enforce transparency so that users know when they are interacting with AI; embed accountability mechanisms that trace and deter misuse; fund public-education campaigns to inoculate people against algorithmic manipulation and impose legal penalties for exploitative applications that verge on criminal.
  
At the same time, developers and open-source communities must bake ethics by default into every release, publishing reference implementations that honor privacy, require consent and preserve user agency. Above all, we must confront in boardrooms, dorm rooms and living rooms alike, how our ethical compass has tilted towards spectacle, how we have normalized feeding the outrage machine, trading empathy for clicks, and elevating fraud, harassment and deceit into must-share content. 
  
We must reclaim an ethic of respect by re-centering consent, curiosity and critical thinking so that even as our tools can mimic any face and voice, we preserve genuine human dignity. In the age of democratized intelligence, the true frontier is not just faster chips or larger datasets but our own maturity, for if we do not evolve as quickly as our code, we risk losing, not technology but our very trust in one another.
 • Suleiman Ayobami Odetoro is An Artificial Intelligence/Machine Learning Professional