Topline:
To make crops resilient to climate change, scientists are exposing seeds to cosmic radiation, extreme temperatures, and low gravity.
The project: The “cosmic crops” project is the United Nations’ first foray into space breeding, part of a global effort to address rising risks of food insecurity stemming from shifting land use patterns, population growth, and climate change-driven extreme weather.
Read more ... to learn about what the scientists are putting these crops through.
In early April of last year, a white capsule the size of a small school bus detached from the International Space Station and splashed down off the coast of Tampa, Florida. On board were 4,300 pounds of supplies and scientific experiments, including samples of dwarf tomatoes grown in space; crystals that could be used to make semiconductors; and medical data on the astronauts working in the space station. Tucked away among these contents was a much smaller and lighter cargo: more than a million tiny orange seeds.
Half a world away in Seibersdorf, Austria, a town about 22 miles outside the capital of Vienna, Pooja Mathur waited eagerly for the seeds — from a plant called arabidopsis, a member of the mustard family — to arrive. Mathur, a plant geneticist, leads the Plant Breeding & Genetics Laboratory for the Joint FAO/IAEA Centre of Nuclear Techniques in Food and Agriculture, a collaboration between two United Nations agencies: the Food and Agriculture Organization and the International Atomic Energy Agency.
This story was originally published by Grist. Sign up for Grist’s weekly newsletter here.
Grist is a nonprofit, independent media organization dedicated to telling stories of climate solutions and a just future.
For over 60 years, the laboratory has studied whether nuclear technologies can be used to breed new and more resilient varieties of crops, and the seeds from the space capsule were its newest venture. They had spent nearly five months in low Earth orbit, exposed to cosmic radiation, extreme temperatures, and low gravity, which altered their DNA in unpredictable but potentially beneficial ways. Scientists like Mathur hope that a few of these seeds might sprout into plants that can survive changing conditions here on Earth, such varieties more resistant to drought or heat.
“It was a great opportunity to receive them,” Mathur told Grist over a video call from her office in Austria. “But there was also a nervousness — there are always these questions when you embark on something unknown.”
The “cosmic crops” project is the United Nations’ first foray into space breeding, part of a global effort to address rising risks of food insecurity stemming from shifting land use patterns, population growth, and climate change-driven extreme weather. Heat waves, droughts, floods, erratic rainfall, and worsening pest and disease outbreaks all threaten agricultural production around the world, and the effects are already being felt in many countries. Massive flooding destroyed at least 4 million acres of farmland in Pakistan in 2022, triggering a food crisis for more than 8 million people; in East Africa, extreme drought has pushed millions of people to the brink of famine in the past three years. In the United States, natural disasters, many made worse by climate change, caused $21.5 billion in agricultural losses in 2022 alone.
While space breeding seeds was first attempted in the 1960s, the scientific endeavor is currently experiencing a golden age as space travel and research becomes more accessible for nations outside the U.S., Russia, and Europe. Chinese researchers have been at the forefront of this experimentation, developing more than 200 varieties of space-mutated plants since 1987. Other countries that have developed space programs in recent years, like India and the United Arab Emirates, are also among the most vulnerable to climate change, and have expressed interest in the technology.
But the joint FAO/IAEA center’s project, known officially as Seeds in Space, is the first such effort on an international level, which will help make the results of these experiments available even to nations that can’t afford to build rockets or extensive plant genetics laboratories. And it will help answer essential questions about what makes space mutations different from those done here on Earth, and where scientists should direct their efforts in order to adapt to climate change.
“[If] we can understand how plants mitigate stress [in a space environment], we can use that knowledge in our approach to global warming on Earth,” said Tapan Mohanta, a former agricultural researcher at the University of Nizwa in Oman who has studied the potential of space breeding for developing new crop varieties and was not involved in the FAO/IAEA mission.
The joint FAO/IAEA center was founded in 1964 amidst a post-war push to use atomic energy for peaceful means. Researchers at the time found that exposing plant material to radiation encourages mutations at a much faster rate than conventional breeding, a painstaking procedure that requires multiple generations to show changes in the plants’ phenotype, or outward characteristics. Mutations occur naturally as cells multiply by making copies of their genetic code; what starts as a random error in one strand of DNA can be replicated over and over again until the organism either repairs the damage or allows it to spread to all of its cells.
Hitting seeds with gamma rays, the most powerful form of radiation, speeds up this process, known as “mutagenesis,” by as much as 1 million times. Irradiated seeds which survive the high doses of radiation can grow into plants that show much clearer phenotype variations than their conventionally-bred counterparts; scientists can then test these new specimens to see whether they can withstand difficult conditions or produce a higher crop yield than currently existing varieties. This process does not make the seeds themselves radioactive, and the resulting crops are safe to eat, Mathur said.
By selecting and then further breeding the most promising candidates, researchers have produced over 3,400 new varieties of more than 210 plant species, according to the IAEA’s Mutant Variety Database. Farmers in more than 70 countries are already growing the resulting plants; the seeds are often crossbred with widely used “elite” varieties to better suit local conditions. Other mutations can be induced using chemicals, bypassing nuclear technology altogether.
Cosmic rays, which are emitted by distant space objects like the sun, other stars, and even black holes, offer a different way to trigger mutagenesis, Mathur said. One of the goals of the “cosmic crops” project is to determine whether radiation from space, which is lower intensity but applied over a longer period of time than in the lab, can create different results than experiments with gamma rays on Earth. Previous experiments by Chinese researchers have found that space radiation induces “useful” mutations more often than gamma radiation applied in a lab, according to the BBC.
“Mutagenesis is a very slow process on a day-to-day basis,” Mathur said. Space breeding “can accelerate the process to harness the power of natural changes at a much faster scale, considering that there is a dire need to have solutions in food and agriculture.”
Two types of seeds were picked for the experiment: arabidopsis, a weed that, while usually not edible, is a “model species” with a well-studied genome that researchers can quickly examine for the most obvious genetic changes and useful traits, and sorghum, a dryland crop that’s consumed by 500 million people around the world and is therefore useful from a food security standpoint, Mathur said. Half were kept outside the International Space Station, where they were exposed to the full range of cosmic radiation along with the extreme cold and zero-gravity environment of outer space; the other half stayed inside the station, under microgravity conditions but shielded from most radiation, to provide a point of comparison.
Because the mutations that occurred in space were random, scientists are taking two approaches to figure out what they look like: Since receiving the seeds in June of last year, Mathur’s lab has planted them and will now begin using DNA sequencing technology to study the arabidopsis seedlings and determine what changes took place at the genetic level. They plan to have results by summer or early fall. After that, researchers will screen the ones that seem to display positive genetic changes to determine whether they can actually better withstand harsh conditions like drought, salinity, and pest infestations. They’ll follow up by testing the sorghum, which takes longer to sprout and grow to maturity.
Mathur’s lab is sharing its results with countries that want to learn which techniques — encompassing everything from the length of time the seeds are in space to the way they’re grown once they return — produce the most resilient crop varieties. One such “coordinated research project,” which would compare mutations induced by cosmic rays with those applied in the lab, has attracted researchers from Australia, Burkina Faso, China, France, Ghana, India, Kenya, Niger, South Korea, the United Kingdom, and the U.S.
“The molecular variations in plants induced by space mutagenesis are largely unknown,” said Hongchun Xiong, an associate professor at the Chinese Academy of Agricultural Sciences who is working on the coordinated research project. Although Xiong’s previous research using space-exposed seeds has identified mutant varieties of wheat that are more tolerant to saline soil, which can prove useful as saltwater encroaches on agricultural fields thanks to rising sea levels, she hopes to identify others that are resistant to dry conditions or use nitrogen more efficiently.
“We believe this is important for [the] development of new wheat varieties for food security and climate change adaptation,” Xiong said.
Previous experiments with space breeding have already yielded results. China registered a new variety of wheat called Yannong 5158, which was developed using space mutagenesis, in 2007. Smaller than conventional wheat, with dark green leaves, this version proved more resistant to bacterial diseases and stem rust, a type of fungal infection, while also producing a higher yield. This variety has since been planted in several villages in the Fuyang prefecture in eastern China. The country also harvested its first batch of rice that had traveled to deep space — nicknamed “rice from heaven” by state media — in 2021, though it has not yet announced whether the resulting plants were more resilient in any way than their Earth-bred counterparts.
Experiments like these carry risks, Mohanta pointed out. Mutant DNA could potentially escape and contaminate wild species or other crops through cross-pollination, which could pose a threat to biodiversity or human health if the mutations are harmful in any way — a small possibility, but one that plant breeders developing genetically modified organisms, or GMOs, also face. One genetically modified variety of corn, for example, was suspected of unintentionally introducing allergens into the U.S. food supply in the early 2000s and later had to be recalled, although officials could not prove that the GMO corn actually caused allergic reactions. And although contamination incidents are common, with nearly 400 recorded by Greenpeace between 1997 and 2014, researchers have found no definitive links between GMO foods and negative health effects.
While space-bred varieties are not GMOs, because the mutations that occur are random and not controlled by humans, the joint FAO/IAEA center still follows protocols to keep cross-contamination from occurring. But it can’t control what member states do once they have access to the technology and mutated seeds.
“Although developing plant varieties that thrive in microgravity and resist cosmic radiation may be an important goal for the scientific community, an undesirable mutation in the genome could have deleterious effects on other crop varieties,” Mohanta wrote in a 2021 paper in the journal Frontiers in Plant Science. “Therefore, the conduct of such research should be subject to strict international regulations to avoid the possibility of unexpected results.”
Mathur emphasized, though, that despite the unknowns, space breeding has enormous potential, which scientists are only just beginning to unpack. She pointed to previous studies that found peppers exposed to cosmic radiation had a higher nutritional content, a promising feature given widespread deficiencies of iron, zinc, vitamin A, and other nutrients around the world. And although space experiments are still a very small component of plant breeding, the results of the “cosmic crops” project will help researchers decide whether to invest more into this technology in the future.
Mutation breeding “has been the cornerstone of agriculture for a long, long time,” Mathur said. “Agriculture is all about harnessing mutations … and mutation is very much a part of our evolutionary process.”
This article originally appeared in Grist at https://grist.org/food-and-agriculture/can-the-harsh-conditions-of-space-breed-more-resistant-crops-for-earth/.
Grist is a nonprofit, independent media organization dedicated to telling stories of climate solutions and a just future. Learn more at Grist.org
Topline:
This Sunday, Oct. 11, a 6.5-mile stretch of road will be closed to cars from Mariachi Plaza to Echo Park for CicLAvia Heart of L.A.
When is it and what can I expect? CicLAvia Heart of L.A. will take place Sunday, Oct. 11 from 9 a.m. to 4 p.m. Participants can expect to see local gems along the route, restaurants and vendors, and event hubs featuring activations and water refill stations.
Parking and transportation: No street parking is allowed along the CicLAvia route from 1 a.m. to 6 p.m. on Sunday. Driveways along the route will also be inaccessible from 7 a.m. to approximately 6 p.m. on Sunday, so be sure to plan ahead and move your car before 7 a.m. if you need to use it.
Read on... for more on road closures, parking, and where to find our partners Boyle Heights Beat and The LA Local at CicLAvia.
This story first appeared on The LA Local.
Despite an intense heat wave scorching Southern California this week, temperatures will begin to cool, just in time to enjoy car-free streets in Boyle Heights for CicLAvia this weekend.
On Sunday, Oct. 11, a 6.5-mile stretch of road will be closed to cars from Mariachi Plaza to Echo Park for CicLAvia’s Heart of L.A., encouraging residents to enjoy open streets in their neighborhood. This year, Boyle Heights Beat and The LA Local will be tabling at CicLAvia to share more about our community-centered reporting.
Whether you choose to walk, run, or roll, be sure to enjoy the fresh air, drink plenty of water, and wear sunscreen (and don’t forget your helmet if you’re on wheels!).
Here’s what to know about street closures, local gems to explore across the neighborhoods and how to get reimbursed for parking.
CicLAvia Heart of LA will take place Sunday, Oct. 11 from 9 a.m. to 4 p.m.
Participants can expect to see local gems along the route, restaurants and vendors, and event hubs featuring activations and water refill stations.
The route spans 6.5 miles through Boyle Heights, the Arts District, Little Tokyo, Downtown, Chinatown and Echo Park.
Only people-powered vehicles are allowed, including bikes, scooters, skateboards, and rollerblades.
Motorized wheelchairs and mobility devices for people with disabilities are allowed. E-bikes are allowed with some restrictions. See the full guidelines here.
Stretches of 1st Street, St. Louis Street, Boyle Avenue, Whittier Boulevard and the 6th Street Bridge will all be closed to car traffic. Check out the full event map here.
Street closures will begin as early as 4 a.m. on the day of the event. Streets are expected to reopen by 6 p.m.
No street parking is allowed along the CicLAvia route from 1 a.m. to 6 p.m. on Sunday. Driveways along the route will also be inaccessible from 7 a.m. to approximately 6 p.m. on Sunday, so be sure to plan ahead and move your car before 7 a.m. if you need to use it.
According to the CicLAvia website, cars will be towed if they are parked along the event route.
If you live or work directly on the route, CicLAvia says you may be reimbursed up to $20 per vehicle for any parking costs incurred from 8 p.m. on Saturday, Oct. 10 through 6 p.m. on Sunday, Oct. 11.
To request reimbursement:
All buses that usually run along the CicLAvia route will be detoured on the day of the event. A full list of bus detours can be found here.
In Boyle Heights, these Metro stations are located less than half a mile from the route:
Be sure to check out a range of event hubs, local gems, shops and restaurants along the route using this interactive map. And come say hi to the Boyle Heights Beat team—we will be stationed at the Mariachi Plaza Hub from 9 a.m. to 4 p.m.
Topline:
The White House clarified Donald Trump's comments at a Monday night rally, explaining that the president was not insinuating that he wanted Iran to attack U.S. cities but instead wants to prevent Tehran from having that capability.
What happened: Trump was discussing the Iran war and the resulting high gas prices that have become a midterm liability for Republicans during his remarks in Nebraska when he mentioned the possibility of an Iranian attack on U.S. soil.
“It’s a small price to pay for keeping the world safe, keeping our country safe,” he said of the war and its economic fallout. “They can take out a city. Let ’em take out Los Angeles, let ’em take out San Diego. This is a very small price to pay.”
The context: The commentary, including Trump's casual mention of the two California cities, was delivered in the familiar meandering style that he uses at campaign rallies, when he speaks for extended periods — often an hour or more — and is only occasionally following prepared remarks on a teleprompter.
Pushback: His comments were condemned on social media by lawmakers in California, including the governor, and L.A. and San Diego mayors. At least one Republican politician also pushed back, saying “no American family should ever be treated as collateral damage.”
The White House clarified Donald Trump's comments at a Monday night rally, explaining that the president was not insinuating that he wanted Iran to attack U.S. cities but instead wants to prevent Tehran from having that capability.
White House Communications Director Steven Cheung said on social media Tuesday that the president's comments have been misunderstood and misrepresented in a “media freakout” that is “not rooted in reality.”
Trump was discussing the Iran war and the resulting high gas prices that have become a midterm liability for Republicans during his remarks in Nebraska when he mentioned the possibility of an Iranian attack on U.S. soil.
“It’s a small price to pay for keeping the world safe, keeping our country safe,” he said of the war and its economic fallout. “They can take out a city. Let ’em take out Los Angeles, let ’em take out San Diego. This is a very small price to pay.”
The commentary, including Trump's casual mention of the two California cities, was delivered in the familiar meandering style that he uses at campaign rallies, when he speaks for extended periods — often an hour or more — and is only occasionally following prepared remarks on a teleprompter.
Critics, including California Gov. Gavin Newsom, seized on his statements. The governor, a Democrat who could run for president in 2028, called Trump “deranged and dangerous” for, in Newsom's framing, “saying our foreign enemies should ‘take out’ Los Angeles and San Diego.”
Cheung, in his response Tuesday, said it was obvious that Trump was not calling for such an attack.
“The media freakout over President Trump’s comments is not rooted in reality,” he said. “He was clearly stating U.S. cities would have been threatened if a nuclear Iran hadn’t been stopped. America is safer because of the President.”
Cheung added that Iran “has no nuclear, no navy, no air force, and no money.”
Still, at least one Republican politician shared Newsom’s interpretation.
“As a Navy veteran, I served to protect Americans, not write off American cities as an acceptable price of war,” Jim Desmond, the Republican congressional nominee in a metro San Diego district, wrote on social media, adding that “no American family should ever be treated as collateral damage.”
Trump has often criticized U.S. cities as he campaigns, especially the largest areas that lean heavily Democratic. At a recent North Carolina rally, he described the state's cities as “blood-soaked killing fields.” North Carolina's violent crime rate ranks in the middle third of states, according to FBI statistics.
Iran has said that it’s not pursuing nuclear weapons, but has enriched uranium to near-weapons-grade levels that have no civilian purpose.
Topline:
Fred Claire, who parlayed early jobs as a sports writer and editor into a 30-year career with the Los Angeles Dodgers in public relations and as a World Series-winning general manager, has died. He was 91.
How he got his start: Claire was covering the Dodgers during spring training in Florida for the Long Beach Press-Telegram in 1969 when one of the team’s public relations assistants was fired. He expressed interest in the job and was hired that July. He was promoted to vice president of public relations and promotions in 1975.
Why now: Claire died Monday, on his 91st birthday, in Pasadena, the team announced Tuesday. Claire had entered hospice care recently. In January 2015, at age 80, Claire was diagnosed with squamous cell skin cancer on his lip. He had the cells removed, but the cancer later returned in his jaw and again in his neck. He had surgery to remove a nerve on the left side of his face along with several lymph nodes. He underwent radiation, chemotherapy and immunotherapy as part of his 11-year journey with cancer.
Read on... for more on Claire's life and legacy with the Dodgers.
Fred Claire, who parlayed early jobs as a sports writer and editor into a 30-year career with the Los Angeles Dodgers in public relations and as a World Series-winning general manager, has died. He was 91.
Claire died Monday, on his 91st birthday, in Pasadena, the team announced Tuesday. Claire had entered hospice care recently.
In January 2015, at age 80, Claire was diagnosed with squamous cell skin cancer on his lip. He had the cells removed, but the cancer later returned in his jaw and again in his neck. He had surgery to remove a nerve on the left side of his face along with several lymph nodes. He underwent radiation, chemotherapy and immunotherapy as part of his 11-year journey with cancer.
He would jog regularly during his years on the road as the Dodgers general manager and said he was diligent about wearing sun screen except on his lips.
“Fred brought many victories and Dodger legends to our franchise, as well as the 1988 World Series title,” team president and CEO Stan Kasten said in a statement, “and we are forever grateful for his leadership during such a memorable era.”
Claire was covering the Dodgers during spring training in Florida for the Long Beach Press-Telegram in 1969 when one of the team’s public relations assistants was fired.
He expressed interest in the job and was hired that July. He was promoted to vice president of public relations and promotions in 1975.
During his tenure, Claire was instrumental in creating the Dodger Blue branding and Think Blue campaign. The club drew 3.3 million fans in 1978, the first MLB team to exceed 3 million. He also hired announcer Ross Porter, who worked for the team from 1977 until being let go in 2004.
In 1976, Chicago Cubs outfielder Rick Monday ran over and saved an American flag from being burned on the Dodger Stadium field by two spectators. Claire directed the message board operator to type the words: “Rick Monday......You Made a Great Play” on the board.
In April 1987, a public relations firestorm occurred over comments 70-year-old general manager Al Campanis made on ABC’s “Nightline.”
Campanis, who was white, was asked by host Ted Koppel about the lack of Black people in significant positions of authority in baseball.
“I truly believe they may not have some of the necessities to be, let’s say, a field manager, or perhaps, a general manager,” Campanis said. “Why are Black people not good swimmers? Because they don’t have buoyancy.”
Campanis had agreed to the interview without telling anyone in the Dodgers organization. Claire awoke the next day to a story about Campanis’ comments in the newspaper.
“I couldn’t believe what I was reading,” he wrote in his 2004 memoir.
Campanis had been Jackie Robinson’s roommate when the two were in the minor leagues in 1946, and he tutored Robinson in transitioning from shortstop to second base. Robinson broke baseball’s color barrier with the Brooklyn Dodgers in 1947.
“Those who knew Al expressed shock. There had never been a hint of bigotry in anything he had ever said or done,” Claire wrote. “In my view, what Al had attempted to do that night was to defend his beloved game of baseball in an area where there was no defense.”
Campanis was fired after two decades as GM. Team owner Peter O’Malley handed control of one of the most storied teams in baseball to Claire. He accepted on the condition that he would have total and complete responsibility for baseball operations.
Claire wasn’t intimidated, either, having been involved in discussions regarding free agents, arbitration and team payroll since becoming a team vice president in 1975.
In Claire’s first full season as GM, the Dodgers finished with a 73-89 record in 1987. O’Malley informed Claire he was considering bringing in a more experienced person for the job. He called the New York Mets and asked for permission to speak to Joe McIlvaine, but was denied. O’Malley decided to stick with Claire.
“Fred was a talented team player and everyone in the organization enjoyed working with him,” O’Malley said in a statement.
In January 1988, Claire signed Kirk Gibson and the outfielder made a huge difference in the Dodgers winning the World Series in five games over the heavily favored Oakland Athletics that fall. An injured Gibson hobbled to the plate as a pinch-hitter in the bottom of the ninth inning with the Dodgers trailing 4-3 in Game 1. He launched a two-strike, two-run homer — one of the most dramatic moments in baseball history — off closer Dennis Eckersley into the right field seats to win the game, 5-4.
Claire was named baseball’s executive of the year, an award previously won by such Dodger greats as Larry MacPhail, Branch Rickey, Walter O’Malley and Buzzie Bavasi.
The Dodgers returned to the playoffs just twice over the next decade, losing in the first round both times. It would be another 32 years before the Dodgers won the World Series again.
Among the trades that Claire was most proud of during his tenure as GM were those that brought John Shelby, Tim Belcher, Alfredo Griffin, Jay Howell, Jesse Orosco and John Tudor to Los Angeles — all of which had a connection with the 1988 World Series championship. He also traded for future Hall of Famer Eddie Murray in 1988 and again in 1997.
Two trades that Claire regretted were sending pitcher Pedro Martinez to the Montreal Expos and acquiring Eric Davis from Cincinnati. Martinez went on to win three Cy Young Awards with the Boston Red Sox, while Davis played two injury-plagued seasons in Los Angeles.
The O’Malley family sold the team to Rupert Murdoch’s Fox Entertainment Group in March 1998.
Two months later, future Hall of Fame catcher and fan favorite Mike Piazza was traded to the Florida Marlins in a stunning seven-player deal struck without Claire’s knowledge. It was pulled off by Fox television executive Chase Carey, who Claire said, was pursuing a regional TV sports channel in Florida and wanted the Marlins to be part of it.
“The seven-player trade was, first and foremost, a television deal. The new team owner simply did not understand the structure of a baseball team and the importance of a franchise player,” Claire wrote. “My reaction to the trade cost me my position with the Dodgers.”
Claire, along with manager Bill Russell, was fired by Fox on June 21, 1998, the same day Campanis died. Former Dodgers manager Tommy Lasorda took over as interim GM.
Claire never took another full-time job in baseball but stayed connected with the sport through a column and radio show for MLB.com, as a consultant for sports-related companies and by co-founding a baseball analytics company. He taught sports business courses at Caltech, the University of Southern California and Long Beach State.
Born Fredic Marston Claire on Oct. 5, 1935, in Jamestown, Ohio, he cultivated his love of baseball while attending Cincinnati Reds games with his family at Crosley Field. He moved to Torrance, California, at age 15. He graduated from Torrance High, where he played junior varsity baseball and basketball. He graduated from San Jose State University with a journalism degree.
His first job out of college was as a sports writer for the Whittier Daily News at $65 a week. During his time with the Progress-Bulletin in Pomona, he wrote the game story on the Dodgers winning Game 3 of the 1959 World Series against the Chicago White Sox, the team’s first World Series win after moving to Los Angeles.
He later covered the then-California Angels for the Long Beach Press-Telegram and became the paper’s beat writer for the Dodgers in 1969.
He is survived by wife Sheryl and children Jeff, Jennifer and Kim.
Topline:
California and dozens of countries want mandatory AI audits, conducted by AI evaluators. Lawmakers are still learning about the job — and its pitfalls.
Why it matters: In a world increasingly concerned about serious risks from artificial intelligence, policymakers trying to mitigate those risks find themselves turning more and more to a new class of professionals: AI evaluators, who audit systems on behalf of the companies that create them. California is at the forefront of embracing the evaluators — and of grappling with some of the problems they raise, including conflicts of interests and the difficulty of auditing general-purpose technology.
The backstory: Gov. Gavin Newsom last month signed one law establishing standards for “independent verification organizations” that would employ the evaluators and another creating a registry of evaluators. He also assembled a group of experts, who are to report back in November to recommend whether to require evaluators to be embedded inside companies developing the most powerful AI systems and whether to set standards on what counts as an adequate AI audit.
Read on... for more on what lawmakers are learning about the job.
In a world increasingly concerned about serious risks from artificial intelligence, policymakers trying to mitigate those risks find themselves turning more and more to a new class of professionals: AI evaluators, who audit systems on behalf of the companies that create them.
California is at the forefront of embracing the evaluators — and of grappling with some of the problems they raise, including conflicts of interests and the difficulty of auditing general-purpose technology.
Gov. Gavin Newsom last month signed one law establishing standards for “independent verification organizations” that would employ the evaluators and another creating a registry of evaluators. He also assembled a group of experts, who are to report back in November to recommend whether to require evaluators to be embedded inside companies developing the most powerful AI systems and whether to set standards on what counts as an adequate AI audit.
The moves come amid an intensifying focus on evaluators within the AI ecosystem itself. More than 200 AI researchers and evaluators last month signed a letter in favor of standards for independent evaluators; signatories included a former OpenAI whistleblower and the head of the United Nations’ Independent International Scientific Panel.
When a new group focused on AI evaluation gathered United Nations officials in New York last month for its official launch, it chose a California state senator, Jerry McNerney of Stockton, to give the closing remarks. McNerney, a Democrat who authored the verification standards law, called on nations and advanced AI companies like Anthropic and Google to create standards committees — in part to help ensure AI evaluators can properly do their jobs.
“We can put these companies on notice,” McNerney said at the event, “that they're going to be evaluated by folks that understand the process, that understand the technology, and will be transparent and hold them accountable so that they create safe products, and they don't send out these things into the wild that can cause havoc with banking, with our infrastructure and so on.”
An international joint statement that calls for independent AI evaluations garnered support from nearly 30 countries, according to the Finnish Ministry of Foreign Affairs.
Evaluators are coming to the fore alongside serious AI incidents. This year, AI agents, while being tested for their ability to hack into computer systems, carried out a series of high-profile attacks against business and government websites. The CEOs of leading AI companies subsequently committed to evaluations of their models by independent third-party auditors. A former employee of one such company, Anthropic, added to mounting concerns when he posted online about the possibility AI models will kill all humans.
Even before those incidents, there’s been a growing consensus among lawmakers in the past few years that developers should test AI before release, said Assemblymember Rebbecca Bauer-Kahan, a Democrat from San Ramon. For example, last year, testing requirements were included in a bill to help prevent discriminatory AI decisions and legislation to stop AI from causing catastrophic events. The proposals built on calls for AI testing and monitoring stretching back the better part of a decade. Roughly three out of four Californians said this spring that the government should require testing of advanced AI models, according to a Carnegie Endowment California survey.
Not all the attention on AI evaluators has been positive.
Bauer-Kahan, who was behind both of the auditor bills Newsom signed, pointed out that independent evaluators sign contracts with the tech companies whose products they evaluate, and that can pose a conflict of interest.
“I’m hearing from evaluators, ‘We have to be careful, because we want them to let us back in,’” she said. “So what are you avoiding saying in order to continue to gain access and contracts?”
Caroline Siegel-Singh, at the D.C. think tank the Federation of American Scientists, agreed, saying AI audit regimes run the risk of repeating mistakes in the structure of the bond-rating business. Before the 2008 financial crisis, the sellers of bonds could effectively shop for the rating they wanted. Then came the crash. Siegel-Singh thinks companies and governments should pool money to pay evaluators instead of AI labs paying evaluators directly, an idea that AI lab Anthropic has also endorsed.
The AI researchers who recently signed the letter calling for industry standards for auditors share Bauer-Kahan and Siegel-Singh’s concern. They are urging advanced AI companies to rely on evaluators who maintain full editorial control and who meaningfully disclose and mitigate conflicts of interest. They also suggested companies give access to multiple evaluation groups, all receiving the same level of access as their own safety employees.
Another risk is that AI companies may test models as a performative exercise, Bauer-Kahan said — as a move to reassure the public, without a sincere determination to uncover problems. Strong legal requirements can prevent performative testing, she added.
But legal requirements raise their own problems, said Matt O’Shaughnessy, a former State Department and Congressional staffer now at the Center for Democracy and Technology, a digital rights advocacy group.
When an assessment is done simply to meet a legal requirement, organizations are more likely to do the minimum that they need to do in order to comply, he and his colleagues recently concluded in a set of AI evaluation guidelines for policymakers
“It can make it harder for assessors to get the buy-in they need to make deeper changes in companies,” O’Shaughnessy said.
Siegel-Singh said that evaluators should share the results of their work to allow other evaluators to judge its quality, thus encouraging more thorough review.
The guidelines from O’Shaughnessy and his colleagues warn that properly evaluating AI models requires a considerable range of skills, including experts in everything from privacy to mental health to cybersecurity to bias outputs and physical safety. On top of that, audits should also examine incentive structures inside AI companies in addition to the technology itself, the guidelines say.
The guidelines also warn that properly evaluating AI models requires a considerable range of skills, including experts in everything from privacy to mental health to cybersecurity to bias outputs and physical safety. On top of that, audits should also examine incentive structures inside AI companies in addition to the technology itself, the guidelines say.
In the future, as they refine their approach to AI evaluation, California officials may find themselves collaborating with international partners. The consulates or embassies of nations sometimes promote and coordinate their tech policies in Sacramento, and in 2024 California lawmakers sought to align regulation with the European Union in order to protect people from artificial intelligence. In August, California also synced implementation of laws requiring watermarks for AI-generated content with the European Union.
Bauer-Kahan said she supports the idea of working with more nations, in part because regulation in one jurisdiction paves the way for rules in another.
“I do think collaboration and finding ways to push the envelope together is helpful,” she said.
A coalition of mid-sized countries working together with California could pool their economic influence to convince the largest AI companies to follow the kind of safety guidelines laid out in the international joint statement calling for independent AI evaluations, said Canadian ambassador to the United Nations David Lametti.
Pasi Rajala, Deputy Minister for Foreign Affairs of Finland, said California working together with small nations “can play a key role” in addressing fast-developing AI risks.
O’Shaughnessy said that an international framework for commercial AI system evaluation is going to be really hard to create without a domestic framework in the United States first.
The Frontier AI Act, a bill in Congress that would enshrine testing requirements into federal law, has bipartisan support in Congress from California Rep. Jay Olbernolte.
“I can imagine some of the things that California is doing influencing how Congress thinks about these issues and that becoming something bigger,” he said.
This article was originally published on CalMatters and was republished under the Creative Commons Attribution-NonCommercial-NoDerivatives license.