Are AI-Generated Search Results Still Protected by Section 230?

Starting this week millions will see AI-generated answers in Google’s search results by default. But the announcement Tuesday at Google’s annual developer conference suggests a future that’s “not without its risks, both to users and to Google itself,” argues the Washington Post:

For years, Google has been shielded for liability for linking users to bad, harmful or illegal information by Section 230 of the Communications Decency Act. But legal experts say that shield probably won’t apply when its AI answers search questions directly. “As we all know, generative AIs hallucinate,” said James Grimmelmann, professor of digital and information law at Cornell Law School and Cornell Tech. “So when Google uses a generative AI to summarize what webpages say, and the AI gets it wrong, Google is now the source of the harmful information,” rather than just the distributor of it…

Adam Thierer, senior fellow at the nonprofit free-market think tank R Street, worries that innovation could be throttled if Congress doesn’t extend Section 230 to cover AI tools. “As AI is integrated into more consumer-facing products, the ambiguity about liability will haunt developers and investors,” he predicted. “It is particularly problematic for small AI firms and open-source AI developers, who could be decimated as frivolous legal claims accumulate.” But John Bergmayer, legal director for the digital rights nonprofit Public Knowledge, said there are real concerns that AI answers could spell doom for many of the publishers and creators that rely on search traffic to survive — and which AI, in turn, relies on for credible information. From that standpoint, he said, a liability regime that incentivizes search engines to continue sending users to third-party websites might be “a really good outcome.”

Meanwhile, some lawmakers are looking to ditch Section 230 altogether. [Last] Sunday, the top Democrat and Republican on the House Energy and Commerce Committee released a draft of a bill that would sunset the statute within 18 months, giving Congress time to craft a new liability framework in its place. In a Wall Street Journal op-ed, Reps. Cathy McMorris Rodgers (R-Wash.) and Frank Pallone Jr. (D-N.J.) argued that the law, which helped pave the way for social media and the modern internet, has “outlived its usefulness.”

The tech industry trade group NetChoice [which includes Google, Meta, X, and Amazon] fired back on Monday that scrapping Section 230 would “decimate small tech” and “discourage free speech online.”
The digital law professor points out Google has traditionally escaped legal liability by attributing its answers to specific sources — but it’s not just Google that has to worry about the issue. The article notes that Microsoft’s Bing search engine also supplies AI-generated answers (from Microsoft’s Copilot). “And Meta recently replaced the search bar in Facebook, Instagram and WhatsApp with its own AI chatbot.”

The article also note sthat several U.S. Congressional committees are considering “a bevy” of AI bills…

Read more of this story at Slashdot.

Utah Locals Are Getting Cheap 10 Gbps Fiber Thanks To Local Governments

Karl Bode writes via Techdirt: Tired of being underserved and overbilled by shitty regional broadband monopolies, back in 2002 a coalition of local Utah governments formed UTOPIA — (the Utah Telecommunication Open Infrastructure Agency). The inter-local agency collaborative venture then set about building an “open access” fiber network that allows any ISP to then come and compete on the shared network. Two decades later and the coalition just announced that 18 different ISPs now compete for Utah resident attention over a network that now covers 21 different Utah cities. In many instances, ISPs on the network are offering symmetrical (uncapped) gigabit fiber for as little as $45 a month (plus $30 network connection fee, so $75). Some ISPs are even offering symmetrical 10 Gbps fiber for around $150 a month: “Sumo Fiber, a veteran member of the UTOPIA Open Access Marketplace, is now offering 10 Gbps symmetrical for $119, plus a $30 UTOPIA Fiber infrastructure fee, bringing the total cost to $149 per month.”

It’s a collaborative hybrid that blurs the line between private companies and government, and it works. And the prices being offered here are significantly less than locals often pay in highly developed tech-centric urban hubs like New York, San Francisco, or Seattle. Yet giant local ISPs like Comcast and Qwest spent decades trying to either sue this network into oblivion, or using their proxy policy orgs (like the “Utah Taxpayer Association”) to falsely claim this effort would end in chaos and inevitable taxpayer tears. Yet miraculously UTOPIA is profitable, and for the last 15 years, every UTOPIA project has been paid for completely through subscriber revenues. […] For years, real world experience and several different studies and reports (including our Copia study on this concept) have made it clear that open access networks and policies result in faster, better, more affordable broadband access. UTOPIA is proving it at scale, but numerous other municipalities have been following suit with the help of COVID relief and infrastructure bill funding.

Read more of this story at Slashdot.

Can Technology Help Reduce Drunk-Driving Deaths?

An anonymous reader shared this report from the Wall Street Journal:
Drunken-driving deaths in the U.S. have risen to levels not seen in nearly two decades, federal data show, a major setback to long-running road-safety efforts. At the same time, arrests for driving under the influence have plummeted, as police grapple with challenges like hiring woes and heightened concern around traffic stops… About 13,500 people died in alcohol impairment-related crashes in 2022, according to data released in April by the National Highway Traffic Safety Administration. That is 33% above 2019’s toll and on par with 2021’s. The last time so many people died as a result of accidents involving intoxicated drivers was in 2006.
That’s still down from the early 1980s, when America was seeing over 20,000 drunk-driving deaths a year, according to the article. “By 2010, that number had fallen to around 10,000 thanks to high-profile public-education campaigns by groups like MADD, tougher laws, and aggressive enforcement that included sobriety checkpoints and typically yielded well over a million DUI arrests annually.”
But some hope to solve the problem using technology:
Many activists and policymakers are banking on the promise of built-in devices to prevent a car from starting if the driver is intoxicated, either by analyzing a driver’s exhaled breath or using skin sensors to gauge the blood-alcohol level. NHTSA issued a notice in December that it said lays the groundwork for potential alcohol-impairment detection technology standards in all new cars “when the technology is mature.”
And Glenn Davis, who manages Colorado’s highway-safety office, “pointed to Colorado’s extensive use of ignition interlock systems that require people convicted of DUI to blow into a tube to verify they are sober in order for their car to start. He said the office promotes nondriving options such as Lyft and Uber.”

Read more of this story at Slashdot.

Why Oregon’s Drug Decriminalization Failed

In 2020 Oregon passed Measure 110, decriminalizing possession of small amounts of drugs.

But now “America’s most radical experiment with drug decriminalization has ended,” writes the Atlantic, “after more than three years of painful results.”

Oregon Governor Tina Kotek has pledged to sign legislation repealing the principal elements of the ballot initiative… Possessing hard drugs is again a crime in Oregon, and courts will return to mandating treatment for offenders. Oregonians had supported Measure 110 with 59 percent of the vote in 2020, but three years later, polling showed that 64 percent wanted some or all of it repealed…

More than $260 million were allocated to services such as naloxone distribution, employment and housing services, and voluntary treatment… Once drugs were decriminalized and destigmatized, the thinking went, those who wanted to continue using would be more willing to access harm-reduction services that helped them use in safer ways. Meanwhile, the many people who wanted to quit using drugs but had been too ashamed or fearful to seek treatment would do so. Advocates foresaw a surge of help-seeking, a reduction in drug-overdose deaths, fewer racial disparities in the health and criminal-justice systems, lower rates of incarceration, and safer neighborhoods for all…

Measure 110 did not reduce Oregon’s drug problems. The drug-overdose-death rate increased by 43 percent in 2021, its first year of implementation — and then kept rising. The latest CDC data show that in the 12 months ending in September 2023, deaths by overdose grew by 41.6 percent, versus 2.1 percent nationwide. No other state saw a higher rise in deaths… Neither did decriminalization produce a flood of help-seeking. The replacement for criminal penalties, a $100 ticket for drug possession with the fine waived if the individual called a toll-free number for a health assessment, with the aim of encouraging treatment, failed completely. More than 95 percent of people ignored the ticket, for which — in keeping with the spirit of Measure 110 — there was no consequence. The cost of the hotline worked out to about $7,000 per completed phone call, according to The Economist. These realities, as well as associated disorder such as open-air drug markets and a sharp rise in violent crime — while such crime was falling nationally — led Oregonians to rethink their drug policy.

The article notes that Oregon was the first U.S. state to decriminalize marijuana back in 1973, and had long shown low rates of imprisonment for non-violent crimes (diverting offenders into so-called “drug courts which could mandate treatment or order court-directed supervision). “However, after Measure 110 was passed and the threat of jail time eliminated, the flow of people into these programs slowed.”

But “One thing Measure 110 got right, at least in principle, is that Oregon’s addiction-treatment system was grossly underfunded,” the article concludes. And it adds that the newly-passed law now “provides extensive new funding for immediate needs, including detox facilities, sobering centers, treatment facilities, and the staff to support those services.”

They recommend other states adopt “adequately funded, evidence-based prevention and treatment” — and instead of punitive incarcerations, “use criminal justice productively to discourage drug use.”

Read more of this story at Slashdot.

FTC and DOJ Think McDonald’s Ice Cream Machines Should Be Legal To Fix

The Federal Trade Commission and the Department of Justice have urged the US Copyright Office to broaden exemptions to the Digital Millennium Copyright Act’s Section 1201. Specifically, the two agencies are advocating for the extension of the right to repair to include “commercial and industrial equipment,” which includes McDonald’s ice cream machines that are notorious for breaking down. The Verge reports: Exemptions to DMCA Section 1201 are issued every three years, as per the Register of Copyrights’ recommendation. Prior exemptions have been issued for jailbreaking cellphones and repairing certain parts of video game consoles. The FTC and DOJ are asking the Copyright Office to go a step further, extending the right to repair to “commercial and industrial equipment.” The comment (PDF) singles out four distinct categories that would benefit from DMCA exemptions: commercial soft serve machines; proprietary diagnostic kits; programmable logic controllers; and enterprise IT. ‘In the Agencies’ view, renewing and expanding repair-related exemptions would promote competition in markets for replacement parts, repair, and maintenance services, as well as facilitate competition in markets for repairable products,” the comment reads.

The inability to do third-party repairs on these products not only limits competition, the agencies say, but also makes repairs more costly and can lead to hundreds or thousands of dollars in lost sales. Certain logic controllers have to be discarded and replaced if they break or if the passwords for them get lost. The average estimated cost of “unplanned manufacturing downtime” was $260,000 per hour, the comment notes, citing research from Public Knowledge and iFixit. As for soft serve machines, breakdowns can lead to $625 in lost sales each day. Business owners can’t legally fix them on their own or hire an independent technician to do so, meaning they have to wait around for an authorized technician — which, the comment says, usually takes around 90 days.

Read more of this story at Slashdot.

Montana’s Governor’s Changes To TikTok Ban Bill Would Ban All Social Media Entirely

Montana Governor Greg Gianforte has returned an “amendatory veto” to the legislature regarding the state’s unconstitutional “ban TikTok” bill, proposing alternative draft language that inadvertently could ban all social media platforms in the state due to poor drafting. The revised language targets any social media application that collects personal information and provides it to a foreign adversary, but since most social media networks collect such information and share it with entities in foreign countries, it would effectively ban all social media in Montana. Techdirt reports: As [1st Amendment lawyer Ari Cohn] points out, the new draft targets any “social media application” that allows for “the collection of personal information or data” and allows for “the personal information or data to be provided to a foreign adversary or a person or entity located within a country designated as a foreign adversary.” Now, some might think that sounds reasonable, but the details here matter. And the details reveal that EVERY social media network collects such information and provides it to people located in countries designated as a foreign adversary. And that’s because “personal information” is a very broad term, as is “provided.” [Ari writes:]

“‘Surely,’ you might think, ‘that just covers the data platforms amass by monitoring and tracking us, right?’ Perhaps not. The bill doesn’t define the term, so who knows what it means in their heads. But we have an idea of what it means out in the real (online) world, by way of the regulations implementing the Children’s Online Privacy Protection Act (COPPA). Those regulations include in the definition of ‘personal information’ things like: First and last name; Online contact information; A screen or user name where it functions in the same manner as online contact information. In other words, the types of information that accompany virtually every piece of content posted on social media. If a platform allows that kind of information to be provided to any foreign adversary or a person or entity located within a foreign adversary, it is banned from Montana.

Do you know who might be persons located within a country designated as a foreign adversary? Users. Users who are provided the kinds of ‘personal information’ that are inherent in the very concept of social media. So, effectively, the bill would ban any social media company that allows any user in China, Russia, Iran, or Cuba to see content from a Montana user (and this is a generous reading, nothing in the bill seems to require that the data/information shared be from a Montana resident). On top of it, each time a user from one of those countries accesses content, platforms would be subject to a $10,000 fine. Do you know which platforms allow people in those countries to access content posted in the United States? All of them. Congratulations, Montana Governor Greg Gianforte. You just managed to accidentally ban all social media for Montanans. Good work.”

Read more of this story at Slashdot.