The UK government doesn't want to sit idle while foundational AI models like ChatGPT flourish. Prime Minister Rishi Sunak and Technology Secretary Michelle Donelan have pledged an initial £100 million (about $124.5 million) to establish a Foundation Model Taskforce. The team will develop AI that ideally makes the country "globally competitive," and will work with the industry to make these systems safer and more reliable.

The taskforce is inspired by the COVID-19 vaccine unit from the height of the pandemic. The group will report directly to both the Prime Minister and Technology Secretary, and have a chairperson announced this summer. The funding comes alongside roughly £900 million ($1.1 billion) in the UK budget devoted to both an exascale supercomputer and dedicated AI research resources.

Officials aren't shy about their hopes. The UK wants to have "sovereign" AI technology that spurs the economy while avoiding the ethical and technical pitfalls that have led experts to call for a six-month pause on experiments. Such models can sometimes be inaccurate or exhibit strange behavior, such as refusing to answer questions or even criticizing users. Donelan sees trustworthy AI as an edge in the field that can help create medical treatments, aid public services and fight climate change.

To some degree, the UK already has a major presence in AI. Google's mainly London-based DeepMind team produces cutting-edge AI research, for instance. However, the hottest systems in recent memory have been developed elsewhere — ChatGPT is the work of US-based OpenAI. The taskforce theoretically keeps British AI relevant despite this trend toward foreign-born technologies.

The controversial EARN IT Act, first introduced in 2020, is returning to Congress after failing twice to land on the president’s desk. The Eliminating Abusive and Rampant Neglect of Interactive Technologies Act, (EARN IT) Act is intended to minimize the proliferation of Child Sexual Abuse Material (CSAM) throughout the web, but detractors say it goes too far and risks further eroding online privacy protections.

Here's how it would work, according to the language of the bill's reintroduction last year. Upon passing, EARN IT would create a national commission composed of politically-appointed law enforcement specialists. This body would be tasked with making a list of best practices to ostensibly curb the digital distribution of CSAM. If online service providers do not abide by these best practices, they would potentially lose blanket immunity under Section 230 of the Communications Decency Act, opening them up to all kinds of legal hurdles — including civil lawsuits and criminal charges.

Detractors say EARN IT places a whole lot of power to regulate the internet in the hands of the commission the bill would create as well as state legislatures. Additionally, language in last year's bill suggests that these guidelines would likely extend to encrypted information, so if an encrypted transmission runs afoul of any guidelines, the platform is on the hook. This will force providers to monitor encrypted communications, which goes against the whole point of encryption in the first place. Additionally, end-to-end encryption is designed so that not even the platform can read the contents. In other words, providers might not be able to offer those protections. 

“This was a dangerous bill two years ago, and because it’s doubled down on its anti-encryption stance, it’s even more dangerous now,” The Center for Internet and Society at Stanford Law School wrote in a blog post last year, a stance also mirrored by the Center for Democracy and Technology. The American Civil Liberties Union, pushing back on a prior version of the bill, said that it "threatens our online speech and privacy rights in ways that will disproportionately harm LGBTQ people, sex workers and others who use the internet to privately communicate and share information and resources."

The Rape, Abuse & Incest National Network (RAINN) has come out in defense of the bill, saying that it will “incentivize technology companies to proactively search for and remove” CSAM materials. “Tech companies have the technology to detect, remove, and stop the distribution of child sexual abuse material. However, there is no incentive to do so because they are subject to no consequences for their inaction,” wrote Erin Earp, RAINN’s interim vice president for public policy.

The bipartisan Senate bills have consistently been introduced by Republican Senator Lindsay Graham and Democrat Senator Richard Blumenthal, and their companion bills in the House likewise have been sponsored by Republican Representative Ann Wagner and Democrat Representative Sylvia Garcia. The full text of H.R.2732 is not publicly available yet, so it's unclear if anything has changed since last year's attempt, though when reintroduced last year it was more of the same. (We've reached out to the offices of Reps. Wagner and Garcia for a copy of the bill's text.) A member of Senator Graham's office confirmed to Engadget that the companion bill will be introduced within the next week. It also remains to be seen if and when this will come up for a vote. Both prior versions of EARN IT died in committee before ever coming to a vote.

In an 89-page complaint unsealed on Monday, the Justice Department alleges 34 current and former members of China’s 912 Special Project Working Group carried out a multi-year campaign to harass critics of Xi Jinping’s regime and discredit American policies. The task force, part of China’s domestic security agency, created thousands of fake social media profiles, including on Twitter and Facebook, to target Chinese dissidents in the US.

In its attempt to disseminate official government propaganda, the group created thousands of fake online personas. Judging from screenshots shared by the Justice Department, many of the profiles did not have more than a dozen accounts following them, but a common thread among them is that they tried to pass as authentic American voices. As The Wall Street Journal points out, one account claimed to be “Susan Miller,” a woman from New York. Another said they were “Julie Torres,” a native of Wisconsin. According to the Justice Department, China’s Ministry of Public Security tracked the performance of the agents involved in the operation and rewarded those who successfully ran multiple online personas without being detected by Twitter and Facebook.

In addition to targeting Chinese dissidents, the group, taking a page from Russia’s disinformation playbook, sought to discredit the US government by exploiting divisions among the American public. For instance, it spread disinformation about George Floyd, the Black Man whose murder by Minneapolis police in 2020 sparked Black Lives Matter protests across the country. The group also amplified Russian propaganda about the war in Ukraine.

“As alleged, the PRC government deploys its national police and the 912 Special Project Working Group not as an instrument to uphold the law and protect public safety, but rather as a troll farm that attacks persons in our country for exercising free speech in a manner that the PRC government finds disagreeable, and also spreads propaganda whose sole purpose is to sow divisions within the United States,” said US attorney Breon Peace, referring to the acronym for the People’s Republic of China.

According to the Justice Department, all 34 of the agents remain at large. This isn’t the first time the US has detailed an effort by China to target overseas dissidents. At the end of last year, US Attorney General Merrick Garland detailed a case involving a multi-year campaign by Chinese operatives to force a US resident to return to China.

Montana is poised to become the first state to ban TikTok. The state’s legislature passed a bill requiring app stores to block the app in the state. The bill will now head to Republican Governor Greg Gianforte, who previously banned the app from state-owned devices.

The ban is slated to go into effect in 2024, though it will likely face legal challenges well before then. Still, the bill has been closely watched as pressure ramps up on TikTok. CEO Shou Chew testified at a Congressional hearing last month where he tried, largely unsuccessfully, to downplay the company’s ties to ByteDance and China.

Meanwhile, the United States government is trying to force ByteDance to divest from TikTok entirely. As The New York Timespoints out, Montana’s ban could serve as a kind of template for the rest of the country, even though it’s unclear whether such bans will hold up to legal challenges.

Like federal lawmakers, the Montana bill claims that TikTok’s ties to ByteDance puts US users personal data at risk because the company could be compelled to turn over information to the Chinese government. TikTok has long denied these claims, and has committed more than a billion dollars to Project Texas, a project meant to address national security concerns.

TikTok has previously said that the proposed ban would hurt small businesses that rely on the app. In a statement following the bill's passage, a TikTok spokesperson called it "egregious government overreach" and said it planned to "fight" the measure. 

“The bill's champions have admitted that they have no feasible plan for operationalizing this attempt to censor American voices and that the bill's constitutionality will be decided by the courts. We will continue to fight for TikTok users and creators in Montana whose livelihoods and First Amendment rights are threatened by this egregious government overreach.”

Developing…

Arkansas has become the second state to pass a law requiring social media platforms to verify the ages of their users. Governor Sarah Huckabee Sanders has signed the “Social Media Safety Act” into law, which requires companies verify the ages of users under 18 and to gain parental consent before minors create new accounts.

Utah recently passed similar measures, which also impose strict rules on how social media platforms handle the accounts of their youngest users, including an age verification requirement. The laws are part of a growing movement to enforce age verification measures and other age-based restrictions for social media companies. At least three other states are currently considering similar legislation.

The laws are also controversial. While lawmakers have for years called for more protections for teens on social media, privacy advocates and civil rights groups have opposed age verification mandates. Likewise, some child safety groups have argued that comprehensive privacy legislation would be a more effective way to protect children.

Under the Arkansas law, slated to take effect in September, companies are required to check the ages of their users via government-issued IDs like a driver’s license or “any commercially reasonable age verification method.” But, as CNNnotes, it’s not entirely clear which companies will be forced to comply with the law. The legislation includes a number of exemptions that would seem to excuse some prominent platforms.

For example, it states that platforms focused on “professional networking” and “career development” are exempt, which would seem to be a direct reference to LinkedIn. It also has exceptions for gaming-focused companies, as well as companies that allow “a user to generate short video clips of dancing, voice overs, or other acts of entertainment in which the primary purpose is not educational or informative,” which could be interpreted as excluding TikTok.

The Biden-Harris administration has proposed stricter pollution standards across the auto industry as part of their goal to increase EV production and slow climate change. The Environmental Protection Agency's (EPA) new guidelines would mean that, by 2032, 67 percent of new light-duty vehicles, such as sedans and light trucks, and 46 percent of new medium-duty vehicles sold will be EVs. The proposed changes could also lead to almost 10 billion tons of carbon dioxide emissions to be avoided through 2055 — about double the US's 2022 emissions. 

Details on the proposed emissions standards aren’t fully clear, but are set to include targets for EV sales and new rules that punish automakers who exceed CO2 standards within the next five to ten years. "Specifically, EPA is proposing stronger CO2 standards for MY 2027 HD vehicles that go beyond the current standards that apply under the HD Phase 2 Greenhouse Gas program," the agency said in a statement. "EPA is also is proposing an additional set of CO2 standards for HD vehicles that would begin to apply in MY 2028, with progressively lower standards each model year through 2032." 

This year has already seen continued progress for EV manufacturing and sales. Tesla reported increased first quarter sales that were 36 percent higher than the same period in 2022, and four percent higher than the last quarter. Volkswagen pledged $193 billion dollars towards EV technology, including batteries and raw materials, over the next five years.

EV purchases are clearly heading in the direction the Biden administration hopes: EV sales accounted for 5.6 percent in 2022, compared to 1.8 percent in 2020. In California, EVs made up almost 19 percent of yearly new car sales. That upswing is set to continue in 2023 with 225,000 EVs sold in the first quarter — about seven percent of all new vehicles. The same period saw used EV purchases increase by almost a third, thanks potentially to their prices decreasing.

At the same time, cost is still a huge barrier for making EVs mainstream. Last year, President Biden signed the Inflation Reduction Act, providing EV buyers subsidies up to $7,500 for SUVs, trucks, and vans under $80,000 and cars under $55,000. Even used EV purchases could get up to $4,000. Recent updated guidance from the US Treasury Department, however, limited how many EVs would qualify for a subsidy. 

Another issue is making sure people have somewhere to charge all these EVs while they're on the go. Currently, there are 130,000 public charging stations across the United States, with President Biden aiming to tackle demand by building another 500,000 by 2030. Part of this plan entails teaming with local officials to build them in widely accessible spaces, with regional governments needing to apply for funds by May 30th. If all goes well, no one would live farther than 50 miles from a charger. 

American officials are taking further steps to set rules for AI systems like ChatGPT. The National Telecommunications and Information Administration (NTIA) is asking for public comments on possible regulations that hold AI creators accountable. The measures will ideally help the Biden administration ensure that these models work as promised "without causing harm," the NTIA says.

While the request is open-ended, the NTIA suggests input on areas like incentives for trustworthy AI, safety testing methods and the amount of data access needed to assess systems. The agency is also wondering if different strategies might be necessary for certain fields, such as healthcare.

Comments are open on the AI accountability measure until June 10th. The NTIA sees rulemaking as potentially vital. There's already a "growing number of incidents" where AI has done damage, the overseer says. Rules could not only prevent repeats of those incidents, but minimize the risks from threats that might only be theoretical.

ChatGPT and similar generative AI models have already been tied to sensitive data leaks and copyright violations, and have prompted fears of automated disinformation and malware campaigns. There are also basic concerns about accuracy and bias. While developers are tackling these issues with more advanced systems, researchers and tech leaders have been worried enough to call for a six-month pause on AI development to improve safety and address ethical questions.

The Biden administration hasn't taken a definitive stance on the risks associated with AI. President Biden discussed the topic with advisors last week, but said it was too soon to know if the technology was dangerous. With the NTIA move, the government is closer to a firm position — whether or not it believes AI is a major problem.

Artificial intelligence has reached a new level of interest ever since ChatGPT burst into the scene. The AI chatbot with its eerily human-like responses has lit a fire under many tech giants and smaller tech companies that are now rushing to release their rival offerings. US President Joe Biden, however, wants them to be careful and make sure that their products are safe before opening them up to the public. According to AP and Reuters, the president has met up with his science and technology advisors, which include academics and executives from Google and Microsoft, to discuss the "risks and opportunities" of artificial intelligence.

While the meeting likely won't culminate in a banning of ChatGPT like what happened in Italy, the president doesn't seem to be convinced that AI is perfectly safe at this point in time. When asked if AI is dangerous, he responded: "It remains to be seen. Could be." He told the group:" "Tech companies have a responsibility, in my view, to make sure their products are safe before making them public...AI can help deal with some very difficult challenges like disease and climate change, but it also has to address the potential risks to our society, to our economy, to our national security."

The White House told the news organizations that the president also used the opportunity to discuss "the importance of protecting rights and safety to ensure responsible innovation and appropriate safeguards." Further, he talked about his previous calls on Congress to pass legislation that would protect kids' privacy online. 

It doesn't sound like policies were made and big changes were planned during the meeting. But Russell Wald from the Stanford Institute for Human-Centered Artificial Intelligence told AP that the president has set "the stage for a national dialogue on the topic by elevating attention to AI." Last year, the Biden administration also released its Blueprint for an AI Bill of Rights. It's meant to guide the design and deployment of AI and other automated systems in a way that protects "the American public in the age of artificial intelligence."

California will require more than half of all heavy trucks sold in the state to be electric by 2035. The rule received approval from the Biden administration today, allowing it to take effect next year, according toThe New York Times. California approved the mandate in 2020 but needed an Environmental Protection Agency (EPA) waiver because it exceeded federal standards.

The rule aims to reduce greenhouse gas emissions from the transportation sector. By 2035, it requires 55 percent of delivery vans and small trucks sold in California to be entirely electric-powered. Similarly, 40 percent of tractor-trailers and 75 percent of buses and larger trucks must be all-electric by the same deadline.

California Governor Gavin Newsom sees the mandate as a bellwether for the nation. “This is a moment to mark because it’s a preview of the order of magnitude of the change in the industry,” Newsom told The New York Times. “There’s a power in these waivers and that power is emulation. We adopt through these waivers the principles and policies that lead to innovation and investment.” Given the size and centrality of California’s economy (it would be the world’s fifth-biggest economy if it were a sovereign nation), the rule would, in practice, essentially apply nationwide — similar to the state’s ban on sales of gas-powered vehicles by 2035.

The trucking industry has criticized the move for its costs and infrastructure requirements. “Drivers don’t want to work in California anymore,” said Jay Grimes, director of federal affairs for the Owner-Operator Independent Drivers Association. “They’re skeptical of the rapid timeline on this transition to electric trucks. Can a trucker get a charge that will take them on a highway for two or three days? Is the technology ready for prime time?” He adds that batteries for electric trucks can weigh thousands of pounds more than combustion engines, potentially limiting hauls. Other truckers have questioned whether the charging station rollout will be adequate for long trips. Finally, electric trucks are more expensive, starting at around $100,000 and stretching into high six figures (although the pricing discrepancies compared to gas trucks could drop over time).

Unsurprisingly, attorneys general from 17 Republican-led states are suing to block the legislation. That list includes (among others) Texas AG Ken Paxton, who has received over $3.9 million in fossil fuel donations since 2002, and Louisiana AG Jeff Landry, who has raked in over $875,000 from oil and gas industries. Their lawsuit is scheduled for the US Court of Appeals for Washington, DC, later this year and could move to the conservative-dominated US Supreme Court afterward.

Clean energy groups acknowledge the mandate’s difficulties but strike an optimistic tone. “There’s a great deal of challenge with the electrification of heavy-duty vehicles,” said Drew Kodjak, executive director of the International Council on Clean Transportation. “But there are elements that lead to optimism.” For example, he points out that government tax incentives and savings from not having to buy gasoline will help with long-term costs. “Companies like FedEx look at the bottom line over the total life span of a vehicle. And when they look long-term, the calculations for this become more optimistic.”

The US Treasury Department issued updated guidance today about which electric vehicles qualify for the federal $7,500 EV tax credit under the Inflation Reduction Act (IRA) that President Biden signed last year. Although the new guidelines add more confusion than clarity, it’s evident that fewer EVs will be eligible.

The updated rules target mineral sourcing in EV batteries, stating that they must be sourced from the US and approved trading partners. That rules out China, which is labeled as a “foreign entity of concern.” Although it’s understandable for the US to limit its dependence on its most powerful adversary, most EVs today run on Chinese-made batteries, making the path forward for receiving the credit on purchases made after April 18th as clear as mud.

To receive tax credits, battery makers must source a significant portion of their materials and manufacturing from North America. Battery components must be 50 percent made or assembled in North America to qualify for a $3,750 credit; critical minerals must be 40 percent sourced from the US or free trade partners for another $3,750 credit. The requirements grow stricter over time, as batteries must be made 100 percent in North America by 2029.

Although some EVs may qualify for partial credits, it’s unclear which models will be eligible after the deadline. “Some EVs will certainly qualify for a partial credit,” said John Bozzella, president and CEO of the Alliance for Automotive Innovation, in a statement to Autoblog. “Given the constraints of the legislation, Treasury's done as well as it could to produce rules that meet the statute and reflect the current market.” However, US officials admit some models will either be reduced or eliminated from the program. The government will publish a revised list of qualifying models by April 18th.

The US and Japan signed a trade agreement on Tuesday that could help long-term by adding the Pacific power to the list of approved partners. In October, the Biden administration announced $2.8 billion in grants for 20 companies to spark domestic EV battery materials and production. The funding, part of the Bipartisan Infrastructure Law, will support the new “American Battery Materials Initiative,” which aims to secure critical EV minerals and boost battery supply to meet Biden’s goal of making EVs half of US vehicle sales by 2030.