On Tuesday, lawyers entered a federal courtroom in Oakland, California, to determine whether Meta should be held accountable for what Facebook and Instagram allegedly did to a generation of children.

That same day, OpenAI introduced a version of ChatGPT specifically designed for teenagers, with stronger parental controls, limits on human-like interactions, additional safety protections, and features intended to steer young users toward learning rather than simply supplying answers.

The coincidence is more than interesting. It may mark the intersection of two technological eras.

Social media is finally being forced to answer hard questions about how its products were designed. Artificial intelligence is arriving just as courts and Congress are beginning to decide whether “we only built the platform” is an adequate defense when foreseeable harm follows.

OpenAI is not on trial in Oakland. But every AI company ought to be watching.

Twenty-nine states sued Meta in federal court in 2023 over child safety and privacy. California, Colorado, Kentucky, and New Jersey are presenting the first consolidated case, which is now underway. The states allege that Meta knowingly designed Facebook and Instagram to keep young people engaged, misrepresented the safety of its products, and collected information from children under 13 without proper parental consent. Meta denies those allegations and says it has invested heavily in protections for younger users.

California Deputy Attorney General Megan O’Neill put the states’ theory plainly in opening arguments. Jurors, she said, would hear that Meta knew a great deal about developing adolescent brains, including children’s sensitivity to rewards and social feedback and their still-maturing impulse control. Meta’s attorney responded that the states are mischaracterizing internal discussions about improving products and overlooking years of safety efforts.

That dispute is for the court to decide. An advisory jury will hear the evidence, but U.S. District Judge Yvonne Gonzalez Rogers will ultimately determine liability.

For the AI industry, however, the questions being asked may prove as important as the eventual verdict.

What did the company know, and when did it know it? That question will anchor six weeks of testimony. Attorneys will press whether internal research flagged the risks to children, whether product features were built to prolong engagement, whether disclosures to parents were honest, and whether children’s data were collected lawfully.

Substitute “AI chatbot” for “social-media platform,” and nearly every question survives.

The technology changes. The accountability question does not.

Artificial intelligence makes those questions harder because the relationship between company and user has grown far more intimate.

Social-media companies learned about us from clicks, likes, searches, purchases, and scrolling behavior. Conversational AI can learn from what we voluntarily tell it. People ask chatbots about illness, marriage problems, finances, work, faith, loneliness, fear, and relationships. Personalized AI systems can remember those conversations and use what they know to shape later responses.

An Axios examination of AI privacy practices this week identified the emerging issue precisely: the question is no longer merely whether companies use conversations to train future models. It is also how information disclosed today may shape an AI system’s memory, personalization, recommendations, advertising, and understanding of the person using it.

Social media learned what kept us clicking. Artificial intelligence can learn why.

That difference matters enormously when the user is a child.

It also explains why the Meta case reaches beyond social media. The legal system is beginning to confront whether technology companies can be held responsible not for content posted by others, but for decisions embedded in their own product design.

That bridge to AI is already appearing elsewhere. In a separate New Mexico case this month, a state court ordered Meta to fund youth-harm remedies and imposed product changes that extend to interactions between minors and Meta’s AI chatbots. Meta is appealing the judgment.

Congress is moving as well.

On August 5, the Senate Commerce Committee advanced a package of child-safety measures that included the Kids Online Safety Act, the Youth AI Privacy Act, the CHATBOT Act, and the Children’s Artificial Intelligence Toy Safety Act.

Those bills move the argument squarely into the AI era. The Youth AI Privacy Act would address advertising to minors, chatbot memory, and design features for young users. The bipartisan CHATBOT Act would give parents greater control over children’s chatbot use and restrict manipulative features.

Meanwhile, OpenAI’s new teen experience limits romantic language and terms of endearment, instructs ChatGPT not to imply that it possesses feelings or consciousness, adds break reminders, and expands parental controls.

None of that establishes wrongdoing by OpenAI or any other company. It establishes something else: policymakers and technology developers are beginning to recognize that minors require a different standard.

The Meta trial could accelerate that change.

Oakland will not write America’s AI code. A single court cannot settle every question surrounding artificial intelligence. Appeals will follow, and Congress still must decide what national rules should govern the technology.

But this case could help establish the legal vocabulary future judges and lawmakers will use: corporate knowledge, foreseeable harm, product design, informed consent, children’s privacy, truthful disclosure, and responsibility for systems deliberately engineered to hold attention.

Those principles could reshape how AI products are tested, how children’s data are retained, how chatbot memory works, and whether companies may profit from advertising informed by deeply personal conversations.

For Christians, there is a larger principle beneath the litigation.

Jesus warned that whoever leads a child astray would be better off drowned at sea with a millstone tied to his neck than to face what awaits him (Matthew 18:6). The legal meaning of corporate liability is one question. The moral principle is clearer.

Power creates responsibility.

A company that holds intimate data on millions of people, designs the system that interacts with them, studies how users respond, and profits from continued use cannot treat that power as an excuse for diminished accountability.

Meta is defending decisions made during the rise of social media. OpenAI, Google, Anthropic, Meta, and their competitors are making decisions now that will define the AI era.

The lesson is not that technological innovation should stop. America needs innovation. But innovation does not suspend responsibility.

Silicon Valley became famous for moving fast and fixing problems later. That approach was always questionable when the product was a website. It becomes far more consequential when the product can learn our confidences, shape our choices, and speak directly to our children.

Meta is on trial in Oakland.

The AI industry should be taking notes.