By Paul Levin

Something important happened in Washington and in a California courtroom this August, and the connection between the two deserves more attention.

On August 5, the Senate Commerce Committee advanced three significant pieces of bipartisan legislation addressing the risks that technology can pose to children: the Kids Online Safety Act, or KOSA; the CHATBOT Act; and the Youth AI Privacy Act. The committee-approved measures seek, in different ways, to address addictive product design, artificial-intelligence chatbots, children's privacy and parental control.

Just three weeks later, Meta reached a historic multistate settlement arising from claims that Facebook and Instagram were deliberately designed in ways that endangered young users.

These events should not be viewed as unrelated developments.

They illustrate something that is frequently overlooked in public debates about lawsuits: civil litigation can do considerably more than compensate someone after an injury has occurred. Properly functioning courts can expose information that otherwise remains private, test corporate conduct against legal duties, create economic consequences for dangerous practices and ultimately help drive broader changes in industry behavior and public policy.

That is particularly important when technology develops faster than legislation.

What happened in Washington

On August 5, 2026, the Senate Committee on Commerce, Science and Transportation advanced KOSA, S.1748; the Youth AI Privacy Act, S.4199; and the CHATBOT Act, S.4407. The latter received unanimous committee support. The committee also advanced the Children's Artificial Intelligence Toy Safety Act.

The proposals attack somewhat different aspects of the same problem.

KOSA would impose obligations on covered platforms concerning design features associated with harms to minors and would provide children and parents greater control over features that can increase or prolong online activity. Its enforcement provisions include actions by the Federal Trade Commission and state attorneys general.

The CHATBOT Act addresses the rapidly emerging world of children interacting with artificial-intelligence systems. Among other provisions, it would require parental involvement in accounts used by children and teenagers and place protective settings around minors' use of AI chatbots. Its introduced version expressly provides that federal law preempts state law only when the two conflict and says states remain free to enact protections greater than those provided federally.

The Youth AI Privacy Act goes further in another important respect. Its introduced text contains an express private right of action allowing a parent or legal guardian to sue over specified violations, with potential remedies including actual damages, punitive damages, attorney's fees and injunctive relief. It likewise preserves state laws that provide greater protections to minors.

The American Association for Justice, on whose Public Affairs Committee I serve, emphasized another important feature of the legislation emerging from committee: preserving families' ability to seek accountability through the courts and protecting stronger state laws rather than allowing federal legislation to erase them.

That issue may sound technical. It is not.

A federal safety standard can become a floor protecting everyone. But if written as an exclusive remedy or accompanied by sweeping preemption, it can also become a ceiling—displacing state consumer-protection laws, common-law duties and remedies that already provide meaningful protection.

Congress should be extremely reluctant to protect children with one hand while inadvertently taking away existing remedies with the other.

Then came Meta

On August 26, Meta reached a sweeping settlement resolving claims brought by 47 states, the District of Columbia, Puerto Rico, American Samoa and the Northern Mariana Islands.

State attorneys general had alleged, among other things, that Meta designed Instagram and Facebook with addictive features, exposed young users to serious risks, improperly collected children's information and misled the public about the safety of its products. Meta did not admit wrongdoing.

The monetary component is extraordinary: up to approximately $17.1 billion under the multistate agreement. But concentrating on the dollars misses much of the significance.

The agreement requires substantial changes directed at the way young people interact with Meta's products. State officials describe measures involving daily-use restrictions, nighttime limits, school-hour protections, age-assurance measures and other changes in the default experience offered to minors.

Whether those reforms ultimately prove adequate remains a legitimate subject of debate. Some critics contend that they leave fundamental aspects of algorithmic engagement untouched. Meta, for its part, has defended its products and its efforts to protect minors.

But something undeniable has happened.

Practices that once seemed largely within the unilateral discretion of enormous technology corporations have become the subject of enforceable obligations negotiated under the pressure of litigation.

That is one of the functions of civil accountability.

The information problem

There is another consequence of litigation that may be still more important than damages.

Lawsuits produce evidence.

Corporations possess enormous amounts of information that consumers, parents, physicians, regulators and even lawmakers may never see. Civil discovery can require the production of emails, research, internal presentations, testing data, safety analyses and communications among executives.

The emerging social-media litigation provides an unusually vivid example.

Earlier this year, a Los Angeles jury found Meta and Google's YouTube negligent in the design or operation of their platforms and found their conduct to be a substantial factor in harming a young plaintiff. The jury awarded $6 million in compensatory and punitive damages. Meta and Google disputed the claims and have challenged the result.

The amount of the verdict is far less significant to the public than what the litigation process exposed.

During that trial, tens of thousands of pages of internal corporate material reportedly became public. Those documents provided an unusual window into what technology companies knew internally about children, engagement, platform design and user well-being.

That information can change the public discussion.

A parent cannot subpoena a multinational technology company.

A journalist ordinarily cannot compel production of its internal research.

A senator cannot necessarily see internal emails merely by asking for them.

A plaintiff who survives the legal requirements of bringing a legitimate lawsuit, however, can use the discovery mechanisms of a court.

That distinction matters enormously.

We have seen this before

The phenomenon is not unique to technology.

Civil litigation has historically operated as an information-producing institution in areas ranging from tobacco and automobiles to pharmaceuticals, medical devices and consumer products.

The sequence is often similar.

A serious injury occurs. Someone begins asking questions. Evidence is gathered. Internal knowledge is compared with external representations. One lawsuit becomes several. Scientists, engineers or physicians examine the evidence. Documents emerge. Courts determine which claims have sufficient support to proceed. Juries sometimes find liability; sometimes they do not. Regulators and legislators obtain a clearer factual record than existed when the controversy began.

Eventually, conduct that was once characterized as an isolated accident may be understood as a systemic safety problem—or, just as importantly, litigation may demonstrate that a proposed theory cannot be proven.

Courts permit both outcomes.

That feature is essential. Civil justice is not supposed to guarantee plaintiffs victory. It provides a structured process in which competing claims can be tested against evidence.

I have experienced that process firsthand in another area involving particularly vulnerable children.

My mass-tort firm, Levin, Rojas, Camassar & Reck, helped initiate the litigation concerning cow's-milk-based nutritional products given to premature infants and the risk of necrotizing enterocolitis, commonly known as NEC.

The firm's first such lawsuit was filed in 2019. Those cases eventually became part of federal multidistrict litigation, MDL No. 3026, in the Northern District of Illinois. Public court records confirm that the litigation concerns allegations that preterm infant nutritional products contributed to NEC, a devastating intestinal disease affecting premature babies.

The defendant vigorously disputes the allegations. That is precisely why there are trials. 

The point is not that litigation establishes truth merely because someone files a complaint. It does not.

The point is that litigation creates a forum in which disputed scientific evidence, corporate knowledge, warnings, medical decisions and causation can be examined under rules of evidence, before an independent judge and ultimately a jury.

That is a very different proposition.

Regulation and litigation are not competitors

There is a persistent tendency in policy debates to present regulation and litigation as competing methods of governing corporate conduct.

That is a false choice.

Government agencies possess expertise and rulemaking authority that private litigants do not. Legislatures can establish prospective rules applicable across entire industries. State attorneys general can protect millions of residents simultaneously. Criminal prosecutors address conduct falling within criminal law.

Private civil litigation does something different.

It begins with an actual person and an actual injury.

It can identify risks regulators have not yet addressed. It can reveal evidence government officials did not possess. And because cases arise from real-world consequences rather than hypothetical rulemaking scenarios, litigation can expose the distance between how a product was supposed to operate and what actually happened to the people using it.

Each institution checks the limitations of the others.

That is why preserving state law and access to courts matters so much as Congress legislates in rapidly changing fields like artificial intelligence and social media.

No federal statute drafted in 2026 can reliably anticipate every way an AI system will interact with children in 2029.

Technology simply moves too quickly.

A strong federal framework therefore should establish meaningful national protections without unnecessarily extinguishing state remedies capable of responding to dangers Congress has not yet imagined.

Accountability can influence design

The Meta settlement demonstrates another feature of litigation that is frequently understated: liability can alter engineering decisions.

The central question in product-safety cases is increasingly not simply what a company said about its product, but how the product itself was designed.

With social media, design choices can include notifications, autoplay, infinite scroll, recommendation systems, visible popularity metrics, beauty filters, default privacy settings and mechanisms intended to increase time spent on a platform.

With artificial intelligence, the analogous choices may involve persistent memory, emotional simulation, unsolicited messages, conversational reinforcement, personalization and the extent to which an AI system encourages a young person to regard it as a human-like confidant.

These are not accidents occurring outside the product.

They are aspects of the product.

The Senate's current proposals reflect growing recognition of that reality. The Youth AI Privacy Act, for example, addresses personalization using minors' personal information, advertising and profiling, while the CHATBOT Act addresses parental consent and family-account controls.

The law is beginning to move upstream—from reacting to harmful content after a child sees it toward examining whether systems themselves were designed with reasonable safeguards.

That is an important evolution.

There are legitimate limits

None of this means every injury should generate liability.

Innovation carries risk. Causation can be extraordinarily difficult to prove. Parents, users, schools and other institutions have responsibilities of their own. Mental-health outcomes in particular can have multiple causes. And courts must distinguish correlation from legal causation.

Technology companies are entitled to defend themselves.

Indeed, robust defense rights are as essential to the legitimacy of civil justice as access to the courthouse is for plaintiffs.

There are also serious policy questions surrounding children's online-safety laws, including privacy, age assurance, parental authority, free expression and the danger that overly broad rules could restrict valuable information.

Those debates should continue.

But they should not be resolved by immunizing an industry from ordinary legal accountability.

The larger lesson

The most important lesson from this month may therefore be institutional rather than technological.

The Senate Commerce Committee is legislating.

State attorneys general are enforcing consumer-protection laws.

Parents and injured young people are bringing civil cases.

Courts are compelling evidence and adjudicating disputed claims.

Juries are evaluating corporate conduct.

And companies are changing products.

That is not evidence of a legal system malfunctioning.

It is evidence of different parts of the American accountability system operating simultaneously.

The cases involving children and technology are still developing. The legislation pending before Congress may change substantially before anything becomes law. Appeals will be filed. Scientific and social-science research will continue. New technologies will create risks that today's lawmakers have not contemplated.

But Congress should preserve one principle as that process unfolds.

When a company creates a product used by millions of children, the people harmed by unreasonable choices should not lose their ability to ask a court to examine what happened.

Regulation can establish the rules.

Enforcement can police them.

But sometimes it takes a lawsuit to uncover why the rules were needed in the first place.