🤖 ARTIFICIAL INTELLIGENCE

States Race To Catch Up With AI

How do you solve a problem like AI?

By taking matters into their own hands, for one. That's what several states including California and Colorado are doing as federal regulation of the technology has been slow-to-non-existent.

“As California has seen with privacy, the federal government isn’t going to act, so we feel that it is critical that we step up in California and protect our own citizens,” Rebecca Bauer-Kahan, a state assembly member told the New York Times. California is pushing some 50 laws on AI that will curb discrimination by AI, copyright infringement and more.

Meanwhile, in Colorado, a law concerning “Consumer Protections in Interactions with Artificial Intelligence" was adopted last month. A similar law was proposed in Connecticut, but ultimately failed. And Utah passed the “Artificial Intelligence Amendments" in, which requires "certain disclosure obligations where a person 'uses, prompts, or otherwise causes generative artificial intelligence to interact with a person',” a white-paper by PerkinsCoie notes.

As the New York Times continues, there are some 400 laws set to regulate AI advancing through state legislatures. This flurry of activity suggests that not only is reigning in AI a pressing issue, but one that enjoys broad support among Americans. So where has Congress been? Apparently far behind: “It’s very hard to do regulations because A.I. is changing too quickly,” Senate Majority Leader Chuck Shumer said. “We didn’t want to rush this.”

"Overall, AI regulations at the state level remain unsettled, with some states pushing forward with regulatory regimes tailored toward their specific concerns," PerkinsCoie explains. "In this context, Colorado's approach to AI regulation stands out as the most comprehensive and risk-based effort in the United States, resembling the E.U. AI Act."

EU AI

While the United States has failed to reach a unified federal policy on AI, the White House did announce an Executive Order in October of 2023 aimed at establishing standards for security and protections for privacy and civil rights.

However, this vacuum in American leadership has led the European Union to step in with the EU AI Act. The act's most influential rules govern how transparent so-called "high-risk" AI systems must be. Furthermore, the law "restricts governments' use of real-time biometric surveillance in public spaces to cases of certain crimes," and bans "the use of artificial intelligence in social scoring, predictive policing and untargeted scraping of facial images from the internet," writes Reuters.

The Verdict

It's clear that failure to regulate AI now—even if that just means safety rails—will mean it's too late in the future. The technology is fast-moving, and could easily evolve beyond our ability to contain it. As such, individual states may prove the necessary stop-gap (or catalyst) to get some guidelines in place now for this technology's development.

📱 SOCIAL MEDIA

Meta Wants To Review Its Reporting

When news broke in 2018 that data firm Cambridge Analytica had wrongfully acquired and misused Facebook user data, share prices of Facebook plummeted. In response, shareholders filed suit against the company, claiming that Facebook did not fully disclose risks to user data and framed such risks as hypothetical, rather than what it already knew of the data breach. However, "the district court dismissed the shareholders’ claims, concluding that they had failed to plead falsity, knowledge of wrongdoing, and loss causation under the elevated standard of Federal Rule of Civil Procedure 9(b), which requires that “a party must state with particularity the circumstances constituting fraud or mistake,” says SCOTUSblog.

But the shareholders appealed, and the US 9th Circuit Court agreed with them—in part—stating that they had plausible claim that Meta (Facebook's parent company) had indeed defrauded them in its presentation of risk factors. "The panel concluded that the shareholders adequately pleaded falsity as to the statements warning that misuse of Facebook users' data could harm Facebook's business, reputation, and competitive position, and the district court erred by dismissing the complaint," the 9th Circuit's panel wrote.

Now, Meta is appeal the 9th Circuit's ruling to the Supreme Court. "[Meta] contends that this case implicates divisions among the federal courts of appeals on two issues: the first on what kinds of risk disclosures public companies must make in their public reports; and the second on whether loss causation allegations are subject to heightened pleading standards under Rule 9(b), or whether ordinary notice pleading under Rule 8 suffices," explains SCOTUSblog.

If the shareholders' case goes to trial, Meta could face a $2 billion settlement, reports Bloomberg.

The Supreme Court will hear the case during it's next term, which begins in October.

Data Breach

Meta's handling of the Cambridge Analytica case isn't the only time it has faced public outcry over data misuse and breaches. In 2021, the personal data of 533 million Facebook users was leaked. As a result, Ireland's Data Protection Commission (DPC), acting under the EU's General Data Protection Regulation (GDPR) laws, fined Meta $276 million. As The Verge notes, the DPC fined Meta some $700 million in 2022 alone.

The Verdict

Sure, its nearly impossible for a company as large as Meta to tell shareholders about every possible risk its users face, but to sit on data misuse as jarring as the Cambridge Analytica case, might not have been the best move either. It will be interesting to see how SCOTUS interprets a firm's reporting responsibilities.