With the hype around AI reaching a fever pitch in recent months, many people fear programs like ChatGPT will one day put them out of a job. For one New York lawyer, that nightmare could become a reality sooner than expected, but not for the reasons you might think. As reported by The New York Times, attorney Steven Schwartz of the law firm Levidow, Levidow and Oberman recently turned to OpenAI’s chatbot for assistance with writing a legal brief, with predictably disastrous results.

Schwartz’s firm has been suing the Columbian airline Avianca on behalf of Roberto Mata, who claims he was injured on a flight to John F. Kennedy International Airport in New York City. When the airline recently asked a federal judge to dismiss the case, Mata’s lawyers filed a 10-page brief arguing why the suit should proceed. The document cited more than half a dozen court decisions, including “Varghese v. China Southern Airlines,” “Martinez v. Delta Airlines” and “Miller v. United Airlines.” Unfortunately for everyone involved, no one who read the brief could find any of the court decisions cited by Mata’s lawyers. Why? Because ChatGPT fabricated all of them. Oops.

In an affidavit filed on Thursday, Schwartz said he had used the chatbot to “supplement” his research for the case. Schwartz wrote he was "unaware of the possibility that [ChatGPT’s] content could be false.” He even shared screenshots showing that he had asked ChatGPT if the cases it cited were real. The program responded they were, claiming the decisions could be found in “reputable legal databases,” including Westlaw and LexisNexis. 

Schwartz said he “greatly regrets” using ChatGPT “and will never do so in the future without absolute verification of its authenticity.” Whether he has another chance to write a legal brief is up in the air. The judge overseeing the case has ordered a June 8th hearing to discuss potential sanctions for the “unprecedented circumstance” created by Schwartz’s actions.

A US federal court this week gave final approval to the $50 million class-action settlement Apple came to last July resolving claims the company knew about and concealed the unreliable nature of keyboards on MacBook, MacBook Air and MacBook Pro computers released between 2015 and 2019. Per Reuters (via 9to5Mac), Judge Edward Davila on Thursday called the settlement involving Apple’s infamous “butterfly” keyboards “fair, adequate and reasonable.” Under the agreement, MacBook users impacted by the saga will receive settlements between $50 and $395. More than 86,000 claims for class member payments were made before the application deadline last March, Judge Davila wrote in his ruling.

Apple debuted the butterfly keyboard in 2015 with the 12-inch MacBook. At the time, former design chief Jony Ive boasted that the mechanism would allow the company to build ever-slimmer laptops without compromising on stability or typing feel. As Apple re-engineered more of its computers to incorporate the butterfly keyboard, Mac users found the design was susceptible to dust and other debris. The company introduced multiple revisions to make the mechanism more resilient before eventually returning to a more conventional keyboard design with the 16-inch MacBook Pro in late 2019.

Apple won’t have to admit wrongdoing as part of the settlement agreement. Before this week, some members of the class action lawsuit attempted to challenge the deal on the grounds that a proposed $125 payout for one group in the class was not enough, an appeal Judge Davila rejected. “The possibility that a better settlement may have been reached — or that the benefits provided under the settlement will not make class members 'whole' — are insufficient grounds to deny approval,” Davila wrote in his ruling. The judge also rejected a request for compensation from MacBook owners who experienced keyboard failures but did not get their computers serviced by Apple. There’s no word when claimants can expect their payment to be sent out, but the lawyers involved in the case said they “look forward to getting the money out to our clients.”

Google has just been hit with a $32.5 million penalty for infringing on a patent held by Sonos. According to Law360, a California federal jury ordered the fine after determining that Google infringed on a patent Sonos holds relating to grouping speakers so they can play audio at the same time, something the company has been doing for years. 

US District Judge William Alsup had already determined that early version of products like the Chromecast Audio and Google Home infringed on Sonos' patent; the question was whether more recent, revamped products were also infringing on the patent. The jury found in favor of Sonos, but decided a second patent — one that relates to controlling devices via a smartphone or other device — wasn't violated. They said that Sonos hadn't convincingly shown that the Google Home app infringed on that particular patent.

Still, it feels like a big win for Sonos, who originally filed suit against Google all the way back in January of 2020. Specifically, Sonos claimed that Google gained knowledge of the patent through prior collaboration between the two companies, back they collaborated to allow for integration between Sonos's speakers and Google Play Music.

Since then, Google counter-sued Sonos, claiming that Sonos had in fact infringed its own patents around smart speakers. As with any good legal battle, Sonos then expanded its own lawsuit a few months later. More recently, Google sued Sonos in 2022, saying that its new voice assistant infringed on seven patents relating to the Google Assistant. 

Whether today's decision will slow the legal battle between the two companies remains to be seen, though we'd expect the bickering to continue full-throttle in the months to come. There are plenty of suits out there between the companies that aren't yet resolved, and we'd expect Google to appeal this decision as well. We've reached out to both Sonos and Google and will update this story with anything we hear.

Korean companies LG and Hyundai are teaming up to build a new EV battery cell manufacturing plant in the US and have signed a memorandum of understanding to invest $4.3 billion in the project. The companies will each hold a stake of 50 percent in the joint venture, which will start construction on the new plant in the second half of 2023. Their new manufacturing facility will be located in Savannah, Georgia, where Hyundai is also building its first all-EV factory in the US. The battery plant is expected to be operational by 2025 at the earliest. After it starts production at full capacity, it will be able to produce 30GHWh of battery every year, which is enough to support the production of 300,000 electric vehicles.

LG and Hyundai are just the latest companies to invest in US-based battery manufacturing facilities over the past couple of years. Toyota announced in 2021 that it will build a battery plant in the country as part of a $3.4 billion investment, while Ultium Cells (GM's and LG's joint venture) secured a $2.5 billion loan from the Energy Department for the construction of EV battery facilities. More recently, Ford announced that it's spending $3.5 billion to build a lithium iron phosphate battery plant in Michigan. Lithium iron phosphate, which can tolerate more frequent and faster charging, costs less than other battery technologies and could bring down the cost of EVs.

Other companies could follow suit, seeing as the Biden administration is pushing to bring more EV and battery manufacturing to the US. Last year, it launched the American Battery Materials Initiative, which will give 20 companies $2.8 billion in grants in hopes of encouraging manufacturers to start battery production stateside and making sure that the US won't be heavily dependent on "unreliable foreign supply chains."

Hyundai and LG believe that the new facility can help create "a stable supply of batteries in the region" and allow them "to respond fast to the soaring EV demand in the US market." Hyundai Mobis, the automaker's parts and service division, will be assembling battery packs using cells manufactured in the plant. The automaker will then use those packs for Hyundai, Kia and Genesis electric vehicles. 

After previewing new verification features last month, LinkedIn is now rolling them out to give job-seekers confidence that they're dealing with real companies and jobs. At the same time, the work-oriented social media site has introduced warnings for messages that may look like scams. The latter feature arrives amidst a spate of fake accounts on the site, according to LinkedIn's latest transparency report. 

The first type of verification tool is related to job postings, displaying information about the posters and their companies. For instance, it can display verifications for a company page and job poster work email, and whether their government ID was verified by CLEAR, the same company that gets people to the front of security lines and airports and other venues. 

LinkedIn

"When you see verifications on job posts, that means there is information that has been verified as authentic by the job poster, LinkedIn or one of our partners," the company wrote in a blog post. Verifying with CLEAR is free but requires sharing a US phone number and government ID. You can also confirm your employer via a verification code sent to your company email address, and some users can verify both their identity and employer via Microsoft Entra, for companies enrolled in the program. 

LinkedIn pointed out that it recently launched the About This Profile feature to show when a profile was created and last updated, and whether it has a verified phone number and/or work email associated with the account. The aim is to flush out fake accounts, but LinkedIn is also rolling out new messages that warn users about high-risk content. 

LinkedIn

"We now also alert you if messages on LinkedIn include high-risk content that could impact your security. For example, we will alert you if a message asks you to take the conversation to another platform, as that can sometimes be a sign of a scam. If something doesn’t feel right, these warnings will also give you the choice to report the content without letting the sender know," the company wrote. 

The new features arrive in the wake of LinkedIn's latest transparency report, which showed a large increase in scam accounts. Between July and December 2022, the company blocked more than 58 million accounts, up from 22 million in the previous six months. LinkedIn is even seeing profiles with fake photos created by AI, something it wrote about last year. Though it said its new "deep-learning-based model proactively checks profile photo uploads to determine if the image is AI-generated," a recent study identified over 1,000 active profiles using AI-generated photos, The Financial Times reported. 

LinkedIn noted that verifications on job posting have just started rolling out, so while you may not see them yet, the tools will be more prevalent as the company expands access. In the meantime, it recommends you check out its tips on how to spot and avoid suspicious job postings. 

China’s cybersecurity regulator has banned Chinese firms from buying chips from US memory manufacturer Micron Technology. Per Reuters, the Cyberspace Administration of China (CAC) said Sunday it found that the company’s products pose “significant security risks” to critical Chinese information infrastructure, including state-owned banks and telecom operators. The ban comes after China announced a review of Micron imports in late March in a move that was seen at the time as retaliation for sanctions Washington has imposed on Chinese chipmakers in recent years.

Idaho-based Micron is the largest memory manufacturer in the US. The Chinese market accounts for about 10 percent of the firm’s annual revenue, though the majority of companies importing Micron products into China are manufacturers making devices for sale in other parts of the world. According to The Wall Street Journal, the CAC’s ban does not apply to non-Chinese firms in China. “We are evaluating the conclusion and assessing our next steps,” Micron told the outlet. “We look forward to continuing to engage in discussions with Chinese authorities.” The CAC did not say what Micron products would be affected by the ban, nor did it share details on what security concerns it had with the company's chips.

The ban is the latest development in an escalating feud over semiconductor technology between the US and China. In recent months, the Biden administration has moved to restrict its rival's access to advanced chipmaking equipment. In January, US, Dutch and Japanese officials agreed to tighten export controls on lithography machines from ASL, Nikon and Tokyo Electron. As The Journal notes, China has been trying to find ways to hit back at the US. Micron was an easy target given that most Chinese companies can turn to suppliers like South Korea’s SK Hynix to make up for any shortfall left by a ban.

DoorDash has been charging iPhone users more than Android users for identical deliveries, according to a lawsuit seeking class action status. The court documents (PDF, via 9to5Mac) submitted for the case included screenshots showing how iPhone users are charged an extra fee for "expanded range." On the company's website, it said the fee "helps DoorDash preserve [customers'] access to the available merchants farthest from [them]." However, the lawsuit said the fee is tacked onto iPhone users' bills more often than Android users' "likely because studies reveal iPhone users earn more."

In addition, it accuses DoorDash of adding the extended range fee onto the total of DashPass subscribers as a way "to subsidize lost revenues from discounted fees." DashPass is the company's $10-a-month subscription service that delivers orders over $12 for free. A couple of screenshots in the court documents show the extended range fee only being added to the account with DashPass and not to the one without, even though they were identical orders made for the same address. 

"DoorDash uses this deceptive practice to trick consumers into believing Dashers receive the 'delivery-related' fees when, in reality, each and every 'delivery fee' is retained in total by DoorDash," the lawsuit states. Other screenshots also showed orders made from iPhones having bigger base delivery fees than orders submitted from Android devices. 

The lawsuit, filed by Ross Hecox and his minor children in the United States District Court of Maryland, is asking for monetary damages of no less than $1 billion "for all consumers who fell prey to DoorDash's illegal pricing scheme over the past four years."

A DoorDash spokesperson denied the allegations and told Insider in a statement:

"The claims put forward in the amended complaint are baseless and simply without merit. We ensure fees are disclosed throughout the customer experience, including on each restaurant storepage and before checkout. Building this trust is essential, and it's why the majority of delivery orders on our platform are placed by return customers. We will continue to strive to make our platform work even better for customers, and will vigorously fight these allegations."

This isn't the first time the delivery service's business practices have been called into question. In 2020, the company, along with GrubHub, Postmates and Uber Eats, were sued for exploiting their dominant position in restaurant deliveries to impose fees on users even during the pandemic. Chicago sued the company for advertising delivery services from restaurants that never consented to be added to its platform. The attorney general for the District of Columbia also filed a lawsuit against DoorDash, accusing it of using tips to cover part of drivers' base pay instead of adding it on top of what they're supposed to get. DoorDash agreed to pay $2.5 million to settle that lawsuit with Washington, DC. 

San Francisco authorities have opened a new investigation into Twitter after six former senior employees filed a lawsuit against the company, according to AP and the San Francisco Chronicle. The plaintiffs are accusing the company of breaking local and federal laws and of violating building codes in its effort to turn some of the rooms in its headquarters into bedrooms for employees. 

The city's authorities first launched an investigation into the company in December 2022 following a Forbes report that it converted some its conference rooms so its staff would have somewhere to rest. If you'll recall, Elon Musk asked remaining employees after a series of mass layoffs to commit to an "extremely hardcore" Twitter that expects them to work "long hours at high intensity." The employees were reportedly given no context about the bedrooms. But one could come to the conclusion, based on his ultimatum, that Musk expected employees to work very long hours that they'd need somewhere to rest in or sleep in overnight. 

One of the plaintiffs in the lawsuit is Joseph Killian, the former lead project manager of global design and construction at Twitter. He said Musk's team instructed him to violate building codes, including removing motion-sensitive lights because they were bother the staff trying to sleep. The company's landlord apparently rejected the request, but he still had to hire an electrician to disconnect the lights without permission. 

Killian also said that he was told to install cheaper locks for the sleeping quarters that don't automatically unlock in case of emergency, even though he warned the team that they'd prevent first responders from accessing the rooms. He reportedly quit that day, though somebody else installed the locks afterward. In addition, Killian accused the company of telling him not to divulge those planned changes to city inspectors visiting Twitter HQ. The inspectors only saw the beds and new furniture and had no idea about the violations, the lawsuit stated. 

Aside from Killian's complaints, the lawsuit also accuses Twitter of not paying their promised severance. The new leadership under Musk, it said, "deliberately, specifically, and repeatedly announced their intentions to breach contracts, violate laws, and otherwise ignore their legal obligations." Regarding Twitter not paying rent, for instance, Musk adviser Pablo Mendoza allegedly told former Twitter real estate division lead Tracy Hawkins: "Elon told me he would only pay rent over his dead body." Alex Spiro, Musk's personal attorney, also allegedly and "loudly opined that it was unreasonable for Twitter’s landlords to expect Twitter to pay rent, since San Francisco was a s—hole." The California Property Trust, which owns the building where Twitter's HQ is located, sued the company for failing to pay $136,250 in rent back in January. 

San Francisco previously gave Twitter 15 days to fix its building permit to be able to keep their beds after Forbes' report came out, but the Chronicle says permits haven't been granted yet. This new investigation is reportedly being conducted by the San Francisco Department of Building Inspection, though it has yet to issue an official statement. As for Twitter, the company hasn't had a communications team in a while. 

AT&T doesn't think that the satellite-to-phone service T-Mobile and SpaceX's Starlink are planning should be approved as it is, and it has informed the FCC of its opinion in a filing (PDF). As Bloomberg notes, the carrier has raised concerns that the companies' service, in its current proposed state, could interfere with existing wireless services. "[I]t is paramount that operations do not jeopardize or inhibit the delivery of terrestrial wireless services," AT&T wrote. 

In August last year, T-Mobile and SpaceX announced a collaboration that would allow the carrier's subscribers to connect to Starlink's second-gen satellites for coverage even if they're in the most remote locations. The companies are gearing up to begin testing this year, and the FCC started seeking comments (PDF) in April on their request to establish supplemental coverage from space (SCS). AT&T's filing is a response to that call. 

The carrier wrote that FCC's rules "do not permit SpaceX’s proposed use of T-Mobile’s terrestrial spectrum" and that the companies "fail to even request — much less justify — rule waivers that would be necessary to authorize their proposed SCS authorizations." It added: "More broadly, the Applicants' technical showings are woefully insufficient regarding the risk of harmful interference posed by their planned SCS deployments. SpaceX and T-Mobile’s applications fall far short of meeting the threshold for waiver and cannot be granted in their current state."

AT&T has plans for a satellite service of its own in partnership with communications specialist AST SpaceMobile. The companies successfully conducted the first two-way satellite audio call on AT&T's network in Texas to a Rakuten number in Japan on a Samsung Galaxy S22 smartphone in April. The carrier assured in its filing that consistent with its comments, "AT&T and AST intend to provide the demonstrations necessary to show that they will not cause interference to any authorized terrestrial system."

Apple is famous for being protective of its projects and for expecting secrecy from its workers. Now, according to The Wall Street Journal, the tech giant is concerned about the possibility of its employees inadvertently leaking proprietary data while using ChatGPT. To prevent that scenario from happening, Apple has reportedly restricted the use of ChatGPT and other AI tools, such as GitHub's Copilot that can autocomplete code. The Journal also says that Apple is working on large language models of its own. 

In early April, The Economist Koreareported that three Samsung employees shared confidential information with ChatGPT. Apparently, one employee asked the chatbot to check database source code for errors, while another asked it to optimize code. The third employee reportedly uploaded a recorded meeting onto the chatbot and asked it to write minutes. It's unclear how Apple is restricting use of generative AI tools and if it's prohibiting their use completely. But in Samsung's case, the company restricted the length of employees' ChatGPT prompts to a kilobyte or 1024 characters of text. 

Large language models like OpenAI's become better the more people use them, because users' interactions are sent back to developers for further training. ChatGPT's terms and conditions, for instance, state that conversations "may be reviewed by [its] AI trainers to improve [its] systems." For a secretive company like Apple, limiting their use doesn't come as a surprise. That said, OpenAI introduced a new privacy control setting in April that enables users to switch off their chat histories so that their conversations can't be used for training. The company made it available after it had to pull ChatGPT for a few hours due to a bug that showed users other people's chat histories. 

Not much is known about Apple's LLM projects at the moment, if there truly are any, but all its AI efforts are under the supervision of John Giannandrea, who used to lead Google's search and AI teams. The tech giant has also snapped up a number of AI startups over the past few years. When asked about AI in an interview recently, Apple chief Tim Cook hinted that the tech giant is taking a cautious approach by saying: "I do think it's very important to be deliberate and thoughtful in how you approach these things."