We live in a paradox. Never before has humanity had access to more information, faster. Yet our decisions, from what we eat to whom we vote for, what we watch and who we date, remain stubbornly resistant to facts alone. Public health campaigns armed with statistics fail to shift behavior. Climate science, however substantive, struggles to ignite action. Heavy economic data rarely changes minds about policy. The uncomfortable truth? We are not the rational creatures we pretend to be.

At every level of human existence, from the individual seeking purpose to communities navigating shared values, from nations crafting policies to our species confronting planetary crises, we operate through a complex interplay of aspirations that give us direction, emotions that move us, thoughts that make sense of chaos, and sensations that anchor us in certainty. Traditional behavior change communication, grounded in knowledge deficit models, treats humans as computers awaiting better data. Generative AI, however, understands what decades of neuroscience have confirmed: we are meaning-making beings first, rational actors second.

The Four Dimensions of Vulnerability

Consider how we actually navigate life. At the individual level, a person doesn’t choose a career based solely on salary projections; they yearn for work that reflects their aspirations, that resonates with who they believe they are becoming. They don’t evaluate political candidates through policy matrices alone; they feel trust or distrust, hope or fear. They don’t absorb information neutrally; they think through narratives that fit their existing worldview, seeking coherence over truth. And they don’t embrace uncertainty willingly; they crave the sensation of solid ground, even if that ground is built on convenient fictions.

Scale this to the collective. Communities form around shared meanings, not shared spreadsheets. Nations rally behind stories, not statistics. Research in behavioral science consistently demonstrates that emotional and identity-based factors overwhelm purely informational interventions. Making people aware of acute meta-level challenges, from climate action, via the pandemic response, to polarization, does rarely fail for lack of information, but flounders due to failures of meaning-making at every nested level of human organization.

This is where generative AI offers something unforeseen: a persuasion technology that speaks fluently in the language of meaning itself, speaking in the alphabet of our emotions.

A Personalized Oracle

Unlike traditional media, which broadcasts the same message to millions, generative AI can craft infinite variations tailored to individual psychologies. It doesn’t just deliver information; it performs empathy.Studies on AI-generated content show that people often perceive it as more emotionally intelligent and persuasive than human-written text, precisely because it can be optimized for resonance rather than constrained by a single author’s perspective.

At the aspiration level, an AI system can present climate action to one person as heroic legacy-building and to another as prudent risk management. At the emotional level, it can deploy hope or fear, belonging or exclusion, tailored to what moves each individual. At the thought level, it constructs narratives that are perceived as internally coherent to each person’s existing mental models, not to change their worldview but to fit snugly within it. At the sensation level, it provides the comforting certainty of answers delivered with confidence, the gravity of a voice that seems to understand our innermost worries.

The implications for misinformation are alarming. Where a conspiracy theory once required human labor to propagate, generative AI can produce thousands of variants, each optimized for different audiences, operating 24/7 at minimal cost. More than spreading lies makes falsifications feel true by wrapping them in the specific meanings that resonate with each recipient.

Scale Without Friction

The economics of persuasion have fundamentally shifted. A political campaign, a corporate lobby, or a state actor can now deploy personalized influence at gigantic scale.Research on AI-mediated communication demonstrates how these systems can identify psychological vulnerabilities, craft messages that exploit them, and iterate based on response, all faster than human fact-checkers can respond.

At the micro level, this means individuals face persuasive messages calibrated precisely to their cognitive biases. At the meso level, communities can be fractured by messages that seem to confirm each subgroup’s existing beliefs while driving wedges between them. At the macro level, nations face information environments where consensus becomes nearly impossible, as each citizen inhabits a personalized reality. At the meta level, coordinated action on planetary challenges becomes harder when the very fabric of shared truth disintegrates.

Formats adapt too. Text for readers, video for visual learners, audio for commuters; interactive conversations that build rapport over time. The interface meets people where they are, in the medium most conducive to acceptance.

The Mirror’s Two Faces

While all the above is worrisome, there is bright potential. The same technology that weaponizes meaning could democratize discernment. Imagine public health communication that speaks to lived experience rather than weaving abstractions, educational content that adapts to learning styles and cultural contexts, and civic engagement tools that coach people to explore complex trade-offs through their own values rather than imposed frameworks.

The challenge resides in our (in)ability to engage with it in a lucid way. We’ve built a persuasion engine more powerful than anything in human history at precisely the moment when our information ecosystem is most fragile, our trust institutions most weakened, our collective challenges most urgent. Building up a stronghold amid the shifting tectonic plates of our hybrid information environment requires combining human literacy – a holistic understanding of self and society, people and planet; and algorithmic literacy, with the acute comprehension of what, why and how our artificial assets operate and impact us.

CHANGE: A Practical Framework

As we navigate this new reality across all dimensions of human existence, consider the CHANGE framework to get started:

**C — Cultivate awareness **of your own vulnerabilities. Recognize that you, too, are susceptible to personalized persuasion. Question why certain messages feel so “right”.

H — Humanize your sources. Seek information from real people with traceable expertise, accountability, and skin in the game, not just compelling AI-generated content.

A — Analyze the medium. Ask not just what is being said but how it’s being delivered. Perfect personalization triggers curiosity, not comfort.

N — Navigate with values. Ground decisions in your core aspirations and ethics, not in the emotional resonance of any single message.

G — Group sense-making. Engage with diverse communities in meaning-making. Collective wisdom across different perspectives resists personalized manipulation.

E — Expect uncertainty. Resist the seductive gravity of easy answers. Truth is often uncomfortable, complex, and incomplete.

Whether the persuasive power of generative AI leads to positive or negative change depends on those who design, deliver and deploy it. Us.

Imagine if AI-powered platforms could serve to make citizens aware of their inherent potential to be agents of change for good. More than ever, society needs people who get proactively involved in the causes that matter to them. From planetary health to children’s rights, via peace efforts to animal protection — the number of reasons to learn about and influence our ever-evolving world is endless. Can we harness AI to wake us up and take part for good before it is too late?