Why We Need a 'Truth Campaign' for the AI Era

Gaurav Laroia and Charlotte Slaiman both previously served at the Federal Trade Commission. They recently published a memo, “Settlement Wins Against Big Tech Should Underwrite Digital Resilience Funds,” at the Federation of American Scientists.

As former attorney-advisors to Federal Trade Commission leadership, we worked hard to turn the page on the era when Big Tech could write off fines for alleged lawbreaking as just the “cost of doing business.” We fought to impose real, substantive limits on corporate data collection and to change the extractive business models fueling digital platforms. But our time in the trenches also taught us a hard truth: while strong injunctive relief and market reforms are vital, enforcement alone isn’t enough.

To truly protect the public, legal and regulatory action should be paired with a massive, proactive public education campaign. Today, as the honeymoon phase for generative AI ends and state attorneys general launch mounting lawsuits against chatbot developers, we have the momentary opportunity to do just that: fund a “Truth Campaign” for the AI era.

The scale of these compounding digital harms resembles an environmental disaster rather than a series of unconnected consumer injuries. Pew polling shows that half of Americans are more concerned than excited about AI, and most doubt their ability to tell whether words or images came from a machine. Gen Z seems to be increasingly anxious about this technology’s impact on their lives. A recent Gallup poll shows their levels of optimism plummeting about how helpful the technology can be in education and especially as deployed in the workforce.

During the social media era, tech companies used algorithmic feeds to hijack our attention and our outrage. Generative AI, with its ability to mimic human communication and create photorealistic images, instead hijacks our trust, imagination, and empathy. As people anthropomorphize chatbots, we are witnessing a rise in parasocial dependency and loneliness and a sharp decline in critical thinking. When a machine uses natural language, our brains are hardwired to assign it empathy and cognition. This illusion of companionship creates a dangerous vulnerability.

Many of these immediate and acute threats are happening to consumers right now. We are seeing a surge in sophisticated, personalized scams driven by technology like voice-cloning and the proliferation of deepfakes used for bullying and harassment. These harms also extend into our physical and civic lives. Flawed facial recognition systems have led to wrongful arrests that disproportionately impact marginalized communities. The unchecked flood of synthetic media threatens to erode baseline public trust in our shared reality. From opaque hiring algorithms to workplace surveillance, these are personal and systemic harms that thrive in an environment of confusion and low digital literacy.

The public is desperately asking for help. It wants clear, unbiased guidance on navigating AI, but right now the balance of messaging is coming from the tech companies themselves. These companies have little financial incentive to tell their users to be skeptical. Chatbots have been marketed as oracles, research assistants, and digital friends. The public needs to know the reality: that these tools are predictive text machines, that they can hallucinate, they are not great at verifying facts, and the intimate information users share with them can be used to train future models.

People deserve concrete solutions. Independent, public interest research and tools could help verify if an image or audio clip is real or synthetic, K-12 training on media skepticism could help protect children from algorithmic manipulation, training for workers to understand their rights if AI is being used for surveillance, or even better funding for the states’ prosecutorial and investigatory apparatus around tech harms could help. But without dedicated funding, who will be able to build and maintain these resources?

We are currently fighting an asymmetric battle on messaging with the companies. They have billions to spend on marketing to spin their narrative on AI. Public interest efforts cannot rely on shoestring budgets to provide the public accurate information. We need a massive and well-resourced investment in the truth to ensure that clear and unbiased guidance reaches the public where they are. We cannot miss this opportunity to build durable civic infrastructure to protect the public. Nor can we afford to subject another generation to the uncontrolled societal experiment that characterized the social media era. Regulators then relied on headline fines that simply disappeared into general government treasuries. Consider the Federal Trade Commission’s historic $5 billion settlement with Facebook in 2019. It was hailed as a historic victory, but as Commissioners Slaughter and Chopra explicitly warned in their dissents, the penalty was unlikely to meaningfully deter the company’s underlying data-harvesting business model. Furthermore, by law, the FTC could not redirect that money to consumer education. But imagine what even a fraction of those funds could have done to educate consumers about their data privacy rights, or about social media’s potential for emotional manipulation?

Fortunately, a new wave of state-level litigation is offering another opportunity to break from the fine-and-forget enforcement model and potentially, put those fines to good use. Two recent cases target the root of these digital harms: defective algorithms and addictive product design. By framing these platforms as defective products engineered to exploit children, enforcers have bypassed the traditional tech liability shield. This breakthrough could open the floodgates for more systemic accountability.

In March, a California jury awarded a 20 year old plaintiff $6 million after finding that Meta and YouTube negligently designed their platforms and caused severe mental health crises under a theory of defective products liability. In the same month a New Mexico jury levied a historic $375 million penalty against Meta for violating the state’s Unfair Practices Act by misleading parents about the safety of their products thereby enabling child exploitation.

These verdicts could be bellwethers for a wave of impending litigation and settlements. Currently, a historic and “sprawling” set of consolidated lawsuits, known as Multidistrict Litigation (MDL) 3047 is proceeding in Federal Court. This lawsuit includes 41 attorneys general, hundreds of school districts, thousands of individual personal injury suits, all consolidated and contesting the “‘unreasonably dangerous’ design of social media platforms.”

These cases, and others, mean billions of dollars may soon be changing hands. That means there’s the rare opportunity to direct a meaningful fraction of these billions towards digital resilience for the public. The critical policy question for state enforcers is whether those funds, after class members and direct victims are made whole through restitution, will disappear into general treasuries or be used to address the real problems confronting the public.

Ideally, governments would fund this education effort directly, but the current state lawsuits provide a rare opening to defend the public now. State Attorneys General and legislatures should be thinking now about how to structure the potential awards and remedies in these cases. As we recently argued, they should direct a portion of these settlement dollars into a Digital Resilience Fund, making sure victims are paid first, then funding the independent watchdogs, media literacy programs, and worker education needed to help people safely live alongside this technology.

Because legal remedies must typically address the specific harms alleged, these dollars can be highly tailored. For instance, if a lawsuit centers on a chatbot endangering minors, the fund could finance school-based AI literacy programs or research into the technology’s psychological impacts. For litigation focusing on deceptive data practices or deepfakes, awards could fund independent tech watchdogs to audit models or launch a campaign to help people spot synthetic media. State attorneys general are constrained in how they can use remedial funds, but state legislatures can step in to supplement these settlements or broaden their mandates. Partnering with them could help transform targeted legal remedies into expansive whole-of-society resilience programs.

We know that this kind of public awareness program can work. When state attorneys general sued the tobacco companies, leading to the 1998 Tobacco Master Settlement Agreement, they used that money to fund the Truth Initiative and education about these products—in perpetuity. By investing in evidence-based culture change around smoking, that campaign, along with new laws, permanently shifted American behavior and changed millions of lives for the better.

The Truth Campaign isn't the only example of successfully increasing societal resilience to pervasive harms. The multi-state opioid settlements mandated that companies pay into abatement funds specifically designed to prevent and treat addiction at the community level. The 2012 national mortgage settlement didn’t just fine the banks; it directed billions towards foreclosure prevention and housing counseling. The 2016 Volkswagen emissions settlement required that the company invest billions into an independent environmental mitigation trust. We can see this educational model working for digital harms, too. A 2023 OECD report shows that when Finland instituted a whole-of-society media literacy program to combat online disinformation it built one of the world's most resilient publics against propaganda. We have the opportunity to build new institutions in the wake of this litigation and we should take it.

As former regulators and advocates on technology policy, we know that protecting the public is not only about guarding against risks. It must also be about equipping ordinary people, parents, the elderly, and kids most of all, with what they are plainly and desperately asking for: the knowledge and means to live with this technology, and to understand what the machine they can talk to in their pocket or at work is and really does.