Science
When A.I.’s Output Is a Threat to A.I. Itself

The internet is becoming awash in words and images generated by artificial intelligence.
Sam Altman, OpenAI’s chief executive, wrote in February that the company generated about 100 billion words per day — a million novels’ worth of text, every day, an unknown share of which finds its way onto the internet.
A.I.-generated text may show up as a restaurant review, a dating profile or a social media post. And it may show up as a news article, too: NewsGuard, a group that tracks online misinformation, recently identified over a thousand websites that churn out error-prone A.I.-generated news articles.
In reality, with no foolproof methods to detect this kind of content, much will simply remain undetected.
All this A.I.-generated information can make it harder for us to know what’s real. And it also poses a problem for A.I. companies. As they trawl the web for new data to train their next models on — an increasingly challenging task — they’re likely to ingest some of their own A.I.-generated content, creating an unintentional feedback loop in which what was once the output from one A.I. becomes the input for another.
In the long run, this cycle may pose a threat to A.I. itself. Research has shown that when generative A.I. is trained on a lot of its own output, it can get a lot worse.
Here’s a simple illustration of what happens when an A.I. system is trained on its own output, over and over again:
This is part of a data set of 60,000 handwritten digits.
When we trained an A.I. to mimic those digits, its output looked like this.
This new set was made by an A.I. trained on the previous A.I.-generated digits. What happens if this process continues?
After 20 generations of training new A.I.s on their predecessors’ output, the digits blur and start to erode.
After 30 generations, they converge into a single shape.
While this is a simplified example, it illustrates a problem on the horizon.
Imagine a medical-advice chatbot that lists fewer diseases that match your symptoms, because it was trained on a narrower spectrum of medical knowledge generated by previous chatbots. Or an A.I. history tutor that ingests A.I.-generated propaganda and can no longer separate fact from fiction.
Just as a copy of a copy can drift away from the original, when generative A.I. is trained on its own content, its output can also drift away from reality, growing further apart from the original data that it was intended to imitate.
In a paper published last month in the journal Nature, a group of researchers in Britain and Canada showed how this process results in a narrower range of A.I. output over time — an early stage of what they called “model collapse.”
The eroding digits we just saw show this collapse. When untethered from human input, the A.I. output dropped in quality (the digits became blurry) and in diversity (they grew similar).
How an A.I. that draws digits “collapses” after being trained on its own output
If only some of the training data were A.I.-generated, the decline would be slower or more subtle. But it would still occur, researchers say, unless the synthetic data was complemented with a lot of new, real data.
Degenerative A.I.
In one example, the researchers trained a large language model on its own sentences over and over again, asking it to complete the same prompt after each round.
When they asked the A.I. to complete a sentence that started with “To cook a turkey for Thanksgiving, you…,” at first, it responded like this:
Even at the outset, the A.I. “hallucinates.” But when the researchers further trained it on its own sentences, it got a lot worse…
An example of text generated by an A.I. model.
After two generations, it started simply printing long lists.
An example of text generated by an A.I. model after being trained on its own sentences for 2 generations.
And after four generations, it began to repeat phrases incoherently.
An example of text generated by an A.I. model after being trained on its own sentences for 4 generations.
“The model becomes poisoned with its own projection of reality,” the researchers wrote of this phenomenon.
This problem isn’t just confined to text. Another team of researchers at Rice University studied what would happen when the kinds of A.I. that generate images are repeatedly trained on their own output — a problem that could already be occurring as A.I.-generated images flood the web.
They found that glitches and image artifacts started to build up in the A.I.’s output, eventually producing distorted images with wrinkled patterns and mangled fingers.
When A.I. image models are trained on their own output, they can produce distorted images, mangled fingers or strange patterns.
A.I.-generated images by Sina Alemohammad and others.
“You’re kind of drifting into parts of the space that are like a no-fly zone,” said Richard Baraniuk, a professor who led the research on A.I. image models.
The researchers found that the only way to stave off this problem was to ensure that the A.I. was also trained on a sufficient supply of new, real data.
While selfies are certainly not in short supply on the internet, there could be categories of images where A.I. output outnumbers genuine data, they said.
For example, A.I.-generated images in the style of van Gogh could outnumber actual photographs of van Gogh paintings in A.I.’s training data, and this may lead to errors and distortions down the road. (Early signs of this problem will be hard to detect because the leading A.I. models are closed to outside scrutiny, the researchers said.)
Why collapse happens
All of these problems arise because A.I.-generated data is often a poor substitute for the real thing.
This is sometimes easy to see, like when chatbots state absurd facts or when A.I.-generated hands have too many fingers.
But the differences that lead to model collapse aren’t necessarily obvious — and they can be difficult to detect.
When generative A.I. is “trained” on vast amounts of data, what’s really happening under the hood is that it is assembling a statistical distribution — a set of probabilities that predicts the next word in a sentence, or the pixels in a picture.
For example, when we trained an A.I. to imitate handwritten digits, its output could be arranged into a statistical distribution that looks like this:
Distribution of A.I.-generated data
Examples of
initial A.I. output:
The distribution shown here is simplified for clarity.
The peak of this bell-shaped curve represents the most probable A.I. output — in this case, the most typical A.I.-generated digits. The tail ends describe output that is less common.
Notice that when the model was trained on human data, it had a healthy spread of possible outputs, which you can see in the width of the curve above.
But after it was trained on its own output, this is what happened to the curve:
Distribution of A.I.-generated data when trained on its own output
It gets taller and narrower. As a result, the model becomes more and more likely to produce a smaller range of output, and the output can drift away from the original data.
Meanwhile, the tail ends of the curve — which contain the rare, unusual or surprising outcomes — fade away.
This is a telltale sign of model collapse: Rare data becomes even rarer.
If this process went unchecked, the curve would eventually become a spike:
Distribution of A.I.-generated data when trained on its own output
This was when all of the digits became identical, and the model completely collapsed.
Why it matters
This doesn’t mean generative A.I. will grind to a halt anytime soon.
The companies that make these tools are aware of these problems, and they will notice if their A.I. systems start to deteriorate in quality.
But it may slow things down. As existing sources of data dry up or become contaminated with A.I. “slop,” researchers say it makes it harder for newcomers to compete.
A.I.-generated words and images are already beginning to flood social media and the wider web. They’re even hiding in some of the data sets used to train A.I., the Rice researchers found.
“The web is becoming increasingly a dangerous place to look for your data,” said Sina Alemohammad, a graduate student at Rice who studied how A.I. contamination affects image models.
Big players will be affected, too. Computer scientists at N.Y.U. found that when there is a lot of A.I.-generated content in the training data, it takes more computing power to train A.I. — which translates into more energy and more money.
“Models won’t scale anymore as they should be scaling,” said Julia Kempe, the N.Y.U. professor who led this work.
The leading A.I. models already cost tens to hundreds of millions of dollars to train, and they consume staggering amounts of energy, so this can be a sizable problem.
‘A hidden danger’
Finally, there’s another threat posed by even the early stages of collapse: an erosion of diversity.
And it’s an outcome that could become more likely as companies try to avoid the glitches and “hallucinations” that often occur with A.I. data.
This is easiest to see when the data matches a form of diversity that we can visually recognize — people’s faces:
This set of A.I. faces was created by the same Rice researchers who produced the distorted faces above. This time, they tweaked the model to avoid visual glitches.
A grid of A.I.-generated faces showing variations in their poses, expressions, ages and races.
This is the output after they trained a new A.I. on the previous set of faces. At first glance, it may seem like the model changes worked: The glitches are gone.
After one generation of training on A.I. output, the A.I.-generated faces appear more similar.
After two generations …
After two generations of training on A.I. output, the A.I.-generated faces are less diverse than the original image.
After three generations …
After three generations of training on A.I. output, the A.I.-generated faces grow more similar.
After four generations, the faces all appeared to converge.
After four generations of training on A.I. output, the A.I.-generated faces appear almost identical.
This drop in diversity is “a hidden danger,” Mr. Alemohammad said. “You might just ignore it and then you don’t understand it until it’s too late.”
Just as with the digits, the changes are clearest when most of the data is A.I.-generated. With a more realistic mix of real and synthetic data, the decline would be more gradual.
But the problem is relevant to the real world, the researchers said, and will inevitably occur unless A.I. companies go out of their way to avoid their own output.
Related research shows that when A.I. language models are trained on their own words, their vocabulary shrinks and their sentences become less varied in their grammatical structure — a loss of “linguistic diversity.”
And studies have found that this process can amplify biases in the data and is more likely to erase data pertaining to minorities.
Ways out
Perhaps the biggest takeaway of this research is that high-quality, diverse data is valuable and hard for computers to emulate.
One solution, then, is for A.I. companies to pay for this data instead of scooping it up from the internet, ensuring both human origin and high quality.
OpenAI and Google have made deals with some publishers or websites to use their data to improve A.I. (The New York Times sued OpenAI and Microsoft last year, alleging copyright infringement. OpenAI and Microsoft say their use of the content is considered fair use under copyright law.)
Better ways to detect A.I. output would also help mitigate these problems.
Google and OpenAI are working on A.I. “watermarking” tools, which introduce hidden patterns that can be used to identify A.I.-generated images and text.
But watermarking text is challenging, researchers say, because these watermarks can’t always be reliably detected and can easily be subverted (they may not survive being translated into another language, for example).
A.I. slop is not the only reason that companies may need to be wary of synthetic data. Another problem is that there are only so many words on the internet.
Some experts estimate that the largest A.I. models have been trained on a few percent of the available pool of text on the internet. They project that these models may run out of public data to sustain their current pace of growth within a decade.
“These models are so enormous that the entire internet of images or conversations is somehow close to being not enough,” Professor Baraniuk said.
To meet their growing data needs, some companies are considering using today’s A.I. models to generate data to train tomorrow’s models. But researchers say this can lead to unintended consequences (such as the drop in quality or diversity that we saw above).
There are certain contexts where synthetic data can help A.I.s learn — for example, when output from a larger A.I. model is used to train a smaller one, or when the correct answer can be verified, like the solution to a math problem or the best strategies in games like chess or Go.
And new research suggests that when humans curate synthetic data (for example, by ranking A.I. answers and choosing the best one), it can alleviate some of the problems of collapse.
Companies are already spending a lot on curating data, Professor Kempe said, and she believes this will become even more important as they learn about the problems of synthetic data.
But for now, there’s no replacement for the real thing.
About the data
To produce the images of A.I.-generated digits, we followed a procedure outlined by researchers. We first trained a type of a neural network known as a variational autoencoder using a standard data set of 60,000 handwritten digits.
We then trained a new neural network using only the A.I.-generated digits produced by the previous neural network, and repeated this process in a loop 30 times.
To create the statistical distributions of A.I. output, we used each generation’s neural network to create 10,000 drawings of digits. We then used the first neural network (the one that was trained on the original handwritten digits) to encode these drawings as a set of numbers, known as a “latent space” encoding. This allowed us to quantitatively compare the output of different generations of neural networks. For simplicity, we used the average value of this latent space encoding to generate the statistical distributions shown in the article.

Science
NIH budget cuts threaten the future of biomedical research — and the young scientists behind it

Over the last several months, a deep sense of unease has settled over laboratories across the United States. Researchers at every stage — from graduate students to senior faculty — have been forced to shelve experiments, rework career plans, and quietly warn each other not to count on long-term funding. Some are even considering leaving the country altogether.
This growing anxiety stems from an abrupt shift in how research is funded — and who, if anyone, will receive support moving forward. As grants are being frozen or rescinded with little warning and layoffs begin to ripple through institutions, scientists have been left to confront a troubling question: Is it still possible to build a future in U.S. science?
On May 2, the White House released its Fiscal Year 2026 Discretionary Budget Request, proposing a nearly $18-billion cut from the National Institutes of Health. This cut, which represents approximately 40% of the NIH’s 2025 budget, is set to take effect on Oct. 1 if adopted by Congress.
“This proposal will have long-term and short-term consequences,” said Stephen Jameson, president of the American Assn. of Immunologists. “Many ongoing research projects will have to stop, clinical trials will have to be halted, and there’ll be the knock-on effects on the trainees who are the next generation of leaders in biomedical research. So I think there’s going to be varied and potentially catastrophic effects, especially on the next generation of our researchers, which in turn will lead to a loss of the status of the U.S. as a leader in biomedical research.“
In the request, the administration justified the move as part of its broader commitment to “restoring accountability, public trust, and transparency at the NIH.” It accused the NIH of engaging in “wasteful spending” and “risky research,” releasing “misleading information,” and promoting “dangerous ideologies that undermine public health.”
National Institutes of Health.
(NIH.gov)
To track the scope of NIH funding cuts, a group of scientists and data analysts launched Grant Watch, an independent project that monitors grant cancellations at the NIH and the National Science Foundation. This database compiles information from public government records, official databases, and direct submissions from affected researchers, grant administrators, and program directors.
As of July 3, Grant Watch reports 4,473 affected NIH grants, totaling more than $10.1 billion in lost or at-risk funding. These include research and training grants, fellowships, infrastructure support, and career development awards — and affect large and small institutions across the country. Research grants were the most heavily affected, accounting for 2,834 of the listed grants, followed by fellowships (473), career development awards (374) and training grants (289).
The NIH plays a foundational role in U.S. research. Its grants support the work of more than 300,000 scientists, technicians and research personnel, across some 2,500 institutions and comprising the vast majority of the nation’s biomedical research workforce. As an example, one study, published in the Proceedings of the National Academy of Sciences, found that funding from the NIH contributed to research associated with every one of the 210 new drugs approved by the U.S. Food and Drug Administration between 2010 and 2016.
Jameson emphasized that these kinds of breakthroughs are made possible only by long-term federal investment in fundamental research. “It’s not just scientists sitting in ivory towers,” he said. “There are enough occasions where [basic research] produces something new and actionable — drugs that will save lives.”
That investment pays off in other ways too. In a 2025 analysis, United for Medical Research, a nonprofit coalition of academic research institutions, patient groups and members of the life sciences industry, found that every dollar the NIH spends generates $2.56 in economic activity.
A ‘brain drain’ on the horizon
Support from the NIH underpins not only research, but also the training pipeline for scientists, physicians and entrepreneurs — the workforce that fuels U.S. leadership in medicine, biotechnology and global health innovation. But continued American preeminence is not a given. Other countries are rapidly expanding their investments in science and research-intensive industries.
If current trends continue, the U.S. risks undergoing a severe “brain drain.” In a March survey conducted by Nature, 75% of U.S. scientists said they were considering looking for jobs abroad, most commonly in Europe and Canada.
This exodus would shrink domestic lab rosters, and could erode the collaborative power and downstream innovation that typically follows discovery. “It’s wonderful that scientists share everything as new discoveries come out,” Jameson said. “But, you tend to work with the people who are nearby. So if there’s a major discovery in another country, they will work with their pharmaceutical companies to develop it, not ours.”
At UCLA, Dr. Antoni Ribas has already started to see the ripple effects. “One of my senior scientists was on the job market,” Ribas said. “She had a couple of offers before the election, and those offers were higher than anything that she’s seen since. What’s being offered to people looking to start their own laboratories and independent research careers is going down — fast.”
In addition, Ribas, who directs the Tumor Immunology Program at the Jonsson Comprehensive Cancer Center, says that academia and industry are now closing their door to young talent. “The cuts in academia will lead to less positions being offered,” Ribas explained. “Institutions are becoming more reluctant to attract new faculty and provide startup packages.” At the same time, he said, the biotech industry is also struggling. “Even companies that were doing well are facing difficulties raising enough money to keep going, so we’re losing even more potential positions for researchers that are finishing their training.”
This comes at a particularly bitter moment. Scientific capabilities are soaring, with new tools allowing researchers to examine single cells in precise detail, probe every gene in the genome, and even trace diseases at the molecular level. “It’s a pity,” Ribas said, “Because we have made demonstrable progress in treating cancer and other diseases. But now we’re seeing this artificial attack being imposed on the whole enterprise.”
Without federal support, he warns, the system begins to collapse. “It’s as if you have a football team, but then you don’t have a football field. We have the people and the ideas, but without the infrastructure — the labs, the funding, the institutional support — we can’t do the research.”
For graduate students and postdoctoral fellows in particular, funding uncertainty has placed them in a precarious position.
“I think everyone is in this constant state of uncertainty,” said Julia Falo, a postdoctoral fellow at UC Berkeley and recording secretary of UAW 4811, the union for workers at the University of California. “We don’t know if our own grants are going to be funded, if our supervisor’s grants are going to be funded, or even if there will be faculty jobs in the next two years.”
She described colleagues who have had funding delayed or withdrawn without warning, sometimes for containing flagged words like “diverse” or “trans-” or even for having any international component.
The stakes are especially high for researchers on visas. As Falo points out for those researchers, “If the grant that is funding your work doesn’t exist anymore, you can be issued a layoff. Depending on your visa, you may have only a few months to find a new job — or leave the country.”
A graduate student at a California university, who requested anonymity due to the potential impact on their own position — which is funded by an NIH grant— echoed those concerns. “I think we’re all a little on edge. We’re all nervous,” they said. “We have to make sure that we’re planning only a year in advance, just so that we can be sure that we’re confident of where that funding is going to come from. In case it all of a sudden gets cut.”
The student said their decision to pursue research was rooted in a desire to study rare diseases often overlooked by industry. After transitioning from a more clinical setting, they were drawn to academia for its ability to fund smaller, higher-impact projects — the kind that might never turn a profit but could still change lives. They hope to one day become a principal investigator, or PI, and lead their own research lab.
Now, that path feels increasingly uncertain. “If things continue the way that they have been,” they said. “I’m concerned about getting or continuing to get NIH funding, especially as a new PI.”
Still, they are staying committed to academic research. “If we all shy off and back down, the people who want this defunded win.”
Rallying behind science
Already, researchers, universities and advocacy groups have been pushing back against the proposed budget cut.
On campuses across the country, students and researchers have organized rallies, marches and letter-writing campaigns to defend federal research funding. “Stand Up for Science” protests have occurred nationwide, and unions like UAW 4811 have mobilized across the UC system to pressure lawmakers and demand support for at-risk researchers. Their efforts have helped prevent additional state-level cuts in California: in June, the Legislature rejected Gov. Gavin Newsom’s proposed $129.7-million reduction to the UC budget.
Earlier this year, a coalition of public health groups, researchers and unions — led by the American Public Health Assn. — sued the NIH and Department of Health and Human Services over the termination of more than a thousand grants. On June 16, U.S. District Judge William Young ruled in their favor, ordering the NIH to reinstate over 900 canceled grants and calling the terminations unlawful and discriminatory. Although the ruling applies only to grants named in the lawsuit, it marks the first major legal setback to the administration’s research funding rollback.
Though much of the current spotlight (including that lawsuit) has focused on biomedical science, the proposed NIH cuts threaten research far beyond immunology or cancer. Fields ranging from mental health to environmental science stand to lose crucial support. And although some grants may be in the process of reinstatement, the damage already done — paused projects, lost jobs and upended career paths — can’t simply be undone with next year’s budget.
And yet, amid the fear and frustration, there’s still resolve. “I’m floored by the fact that the trainees are still devoted,” Jameson said. “They still come in and work hard. They’re still hopeful about the future.”
Science
Should bioplastics be counted as compost? Debate pits farmers against manufacturers
Greg Pryor began composting yard and food waste for San Francisco in 1996, and today he oversees nine industrial-sized composting sites in California and Oregon that turn discarded banana peels, coffee grounds, chicken bones and more into a dark, nutrient-rich soil that farmers covet for their fields and crops.
His company, Recology, processes organic waste from cities and municipalities across the Bay Area, Central Valley, Northern California, Oregon and Washington — part of a growing movement to reduce greenhouse gas emissions by minimizing food waste in landfills.
But, said Pryor, if bioplastic and compostable food packaging manufacturers’ get their way, the whole system could collapse.
At issue is a 2021 California law, known as Assembly Bill 1201, which requires that products labeled “compostable” must actually break down into compost, not contaminate soil or crops with toxic chemicals, and be readily identifiable to both consumers and solid waste facilities.
The law also stipulates that products carrying a “compostable” label must meet the U.S. Department of Agriculture’s National Organic Program requirements, which only allow for plant and animal material in compost feedstock, and bar all synthetic substances and materials — plastics, bioplastics and most packaging materials — except for newspaper or other recycled paper without glossy or colored ink.
Close-up of text on plastic cup reading Made From Corn, referring to plant derived bioplastics.
(Getty Images)
The USDA is reviewing those requirements at the request of a compostable plastics and packaging industry trade group. Its ruling, expected this fall, could open the door for materials such as bioplastic cups, coffee pods and compostable plastic bags to be admitted into the organic compost waste stream.
Amid pressure from the industry, the California Department of Resources Recycling and Recovery said it will await implementing its own rules on AB 1201 — originally set for Jan. 1, 2026 — until June 30, 2027, to incorporate the USDA guidelines, should there be a change.
Pryor is concerned that a USDA ruling to allow certain plastic to be considered compost will contaminate his product, make it unsaleable to farmers, and undermine the purpose of composting — which is to improve soil and crop health.
Plastics, microplastics and toxic chemicals can hurt and kill the microorganisms that make his compost healthy and valued. Research also shows these materials, chemicals and products can threaten the health of crops grown in them.
And while research on new generation plastics made from plant and other organic fibers have more mixed findings — suggesting some fibers, in some circumstances, may not be harmful — Pryor said the farmers who buy his compost don’t want any of it. They’ve told him they won’t buy it if he accepts it in his feedstock.
“If you ask farmers, hey, do you mind plastic in your compost? Every one of them will say no. Nobody wants it,” he said.
However, for manufacturers of next-generation, “compostable” food packaging products — such as bioplastic bags, cups and takeout containers made from corn, kelp or sugarcane fibers — those federal requirements present an existential threat to their industry.
That’s because California is moving toward a new waste management regime which, by 2032, will require all single-use plastic packaging products sold in the state to be either recyclable or compostable.
A worker at Recology’s Blossom Valley composting site rides his bike back to the sorting machines after a break in Vernalis, Calif., on June 26.
(Susanne Rust / Los Angeles Times)
If the products these companies have designed and manufactured for the sole purpose of being incorporated in the compost waste stream are excluded, they will be shut out of the huge California market.
They say their products are biodegradable, contain minimal amounts of toxic chemicals and metals, and provide an alternative to the conventional plastics used to make chip bags, coffee pods and frozen food trays — and wind up in landfills, rivers and oceans.
“As we move forward, not only are you capturing all this material … such as coffee grounds, but there isn’t really another packaging solution in terms of finding an end of life,” for these products, said Alex Truelove, senior policy manager for the Biodegradable Product Institute, a trade organization for compostable packaging producers.

(Robert Gauthier/Los Angeles Times)

Material is loaded into a mixing truck where biosolids and amendments are combined then stored in climate controlled piles to cure at the Tulare Lake Compost plant. (Robert Gauthier/Los Angeles Times)

“Even if you could recycle those little cups, which it seems like no one is willing to do … it still requires someone to separate out and peel off the foil top and dump out the grounds. Imagine if you could just have a really thin covering or really thin packaging, and then you could just put it all in” the compost he said. “How much more likely would it be for people to participate?”
Truelove and Rhodes Yepsen, the executive director of the bioplastic institute, also point to compost bin and can liners, noting that many people won’t participate in separating out their food waste if they can’t put it in a bag — the “yuck” factor. If you create a compostable bag, they say, more people will buy into the program.
The institute — whose board members include or have included representatives from the chemical giant BASF Corp., polystyrene manufacturer Dart Container, Eastman Chemical Co. and PepsiCo — is lobbying the federal and state government to get its products into the compost stream.
Greg Pryor, Recology’s director of landfill and organics, stands in front of a pile of processed compost at the integrated waste management’s Blossom Valley compost site in Vernalis, Calif., on June 26.
(Susanne Rust / Los Angeles Times)
The institute also works as a certifying body, testing, validating and then certifying compostable packaging for composting facilities across the U.S. and Canada.
In 2023, it petitioned the USDA to reconsider its exclusion of certain synthetic products, calling the current requirements outdated and “one of the biggest stumbling blocks” to efforts in states, such as California, that are trying to create a circular economy, in which products are designed and manufactured to be reused, recycled or composted.
In response, the federal agency contracted the nonprofit Organics Material Review Institute to compile a report evaluating the research that’s been conducted on these products’ safety and compostability.
The institute’s report, released in April, highlighted a variety of concerns including the products’ ability to fully biodegrade — potentially leaving microplastics in the soil — as well as their tendency to introduce forever chemicals, such as per- and polyfluoroalkyl substances (PFAS), and other toxic chemicals into the soil.
“Roughly half of all bioplastics produced are non-biodegradable,” the authors wrote. “To compensate for limitations inherent to bioplastic materials, such as brittleness and low gas barrier properties, bioplastics can contain additives such as synthetic polymers, fillers, and plasticizers. The specific types, amounts, and hazards of these chemicals in bioplastics are rarely disclosed.”
The report also notes that while some products may break down relatively efficiently in industrial composting facilities, when left out in the environment, they may not break down at all. What’s more, converting to biodegradable plastics entirely could result in an increase in biodegradable waste in landfills — and with it emissions of methane, a powerful greenhouse gas, the authors wrote.
Yepsen and Truelove say their organization won’t certify any products in which PFAS — a chemical often used to line cups and paper to keep out moisture — was intentionally added, or which is found in levels above a certain threshold. And they require 90% biodegradation of the products they certify.
Judith Enck, a former regional Environmental Protection Agency director, and the founder of Beyond Plastics, an anti-plastic waste environmental group based in Bennington, Vt., said the inclusion of compost as an end-life option for packaging in California’s new waste management regime was a mistake.
“What it did was to turn composting into a waste disposal strategy, not a soil health strategy,” she said. “The whole point of composting is to improve soil health. But I think what’s really driving this debate right now is consumer brand companies who just want the cheapest option to keep producing single-use packaging. And the chemical companies, because they want to keep selling chemicals for packaging and a lot of so-called biodegradable or compostable packaging contains those chemicals.”
Bob Shaffer, an agronomist and coffee farmer in Hawaii, said he’s been watching these products for years, and won’t put any of those materials in his compost.
“Farmers are growing our food, and we’re depending on them. And the soils they grow our crops in need care,” he said. “I’ll grow food for you, and I’ll grow gorgeous food for you, but give us back the food stuff you’re not using or eating, so we can compost it, return it to the soil, and make a beautiful crop for you. But be mindful of what you give back to us. We can’t grow you beautiful food from plastic and toxic chemicals.”
Recology’s Pryor said the food waste his company receives has increasingly become polluted with plastic.
He pointed toward a pile of food waste at his company’s composting site in the San Joaquin Valley town of Vernalis. The pile looked less like a heap of rotting and decaying food than a dirty mound of plastic bags, disposable coffee cups, empty, greasy chip bags and takeout boxes.
“I’ve been doing this for more than three decades, and I can tell you the food we process hasn’t changed over that time,” he said. “Neither have the leaves, brush and yard clippings we bring in. The only thing that’s changed? Plastics and biodegradable plastics.”
He said if the USDA and CalRecycle open the doors for these next-generation materials, the problem is just going to get worse.
“People are already confused about what they can and can’t put in,” he said. “Opening the door for this stuff is jut going to open the floodgates. For all kinds of materials. It’s a shame.”
Science
Federal contractors improperly dumped wildfire-related asbestos waste at L.A. area landfills

Federal contractors tasked with clearing ash and debris from the Eaton and Palisades wildfires improperly sent truckloads of asbestos-tainted waste to nonhazardous landfills, including one where workers were not wearing respiratory protection, according to state and local records.
From Feb. 28 to March 24, federal cleanup crews gathered up wreckage from six burned-down homes as part of the wildfire recovery efforts led by the U.S. Army Corps of Engineers and its primary contractor Environmental Chemical Corp.
However, prior to reviewing mandated tests for asbestos, crews loaded the fire debris onto dump trucks bound for Simi Valley Landfill and Recycling Center, and possibly Calabasas Landfill in unincorporated Agoura and Sunshine Canyon Landfill in Los Angeles’ Sylmar neighborhood, according to reports by the California Office of Emergency Services and Ventura County.
Later on, federal contractors learned those tests determined that the fire debris from these homes contained asbestos, a fire-resistant building material made up of durable thread-like fibers that can cause serious lung damage if inhaled.
The incident wasn’t reported to landfill operators or environmental regulators until weeks later in mid-April.
Many Southern California residents and environmental groups had already objected to sending wildfire ash and debris to local landfills that were not designed to handle high levels of contaminants and potentially hazardous waste that are often commingled in wildfire debris. They feared toxic substances — including lead and asbestos — could pose a risk to municipal landfill workers and might even drift into nearby communities as airborne dust.
The botched asbestos disposal amplifies those concerns and illustrates that in some cases federal contractors are failing to adhere to hazardous waste protocols.
“You have to wonder if they caught it here, how many times didn’t they catch it?” asked Jane Williams, executive director of the nonprofit California Communities Against Toxics. “It’s the continued failure to effectively protect the public from the ash. This is further evidence of that failure. This is us deciding those who work and live around these landfills are expendable.”
As of May 1, nearly 1 million tons of disaster debris has been taken to four landfills in Southern California. Simi Valley, an 887-acre landfill in Ventura County, has taken two-thirds of the tonnage. Several residents who live nearby voiced their disappointment ahead of the June 24 Ventura County Board of Supervisors vote to approve emergency waivers to allow fire debris to continue to be disposed of at Simi Valley Landfill — without a cap on tonnage — until Sept. 3.
“When I told my kids about the fire debris being dumped at the landfill, they asked me, why would anyone allow us to be exposed to this?” said Nicole Luekenga, a resident of nearby Moorpark, at the June 24 board meeting. “We are deeply concerned about the potential health risks from the fire debris being dumped at a residential landfill in our community. It feels as though profit and convenience are being prioritized over public safety, and that is unacceptable.”
An Environmental Chemical Corp. official acknowledged the lapse in asbestos protocols led to the improper disposal in February and March. He said the ash and debris from the six homes — four in Altadena, one in Pacific Palisades and one in Malibu — contained “trace amounts” of asbestos but did not elaborate on the specific type of building material that contained asbestos, or why the debris wasn’t flagged.
Asbestos has historically been used in a variety of construction materials — large and small — including roofing shingles, cement pipes, popcorn ceilings and insulation.
The company official said the improper disposal may have been due to a failure of either its workers or subcontractors to properly review paperwork. He also said he was unaware of any other cases in which asbestos or hazardous waste were improperly disposed. The Army Corps of Engineers declined to comment on the matter.
Environmental Chemical officials told Simi Valley Landfill that the asbestos should be presumed to be friable, a form of the fibrous mineral that is more easily broken down into smaller pieces and considered hazardous waste, according to an April letter from the landfill’s owner, Waste Management, to the Los Angeles Regional Water Quality Control Board.
During the time the asbestos waste was taken to the landfills, workers handling fire debris at Simi Valley Landfill had not been wearing protective masks or respirators, according to inspection reports. Typically municipal landfill workers don’t wear face coverings because they are mostly handling trash and nonhazardous waste.
But experts say protective masks are essential for protecting worker health at landfills. Landfill workers or hired contractors regularly drill pipelines extending hundreds of feet underground into the layers of the waste to extract gases that can build up when garbage decomposes. Experts say drilling into hazardous waste, such as asbestos waste, could expose workers to harmful substances if they aren’t wearing appropriate protective equipment.
During at least one visit in March, a Ventura County inspector found workers without masks in parts of the landfill designated for fire debris. Waste Management staff told the county inspector that mask-wearing was voluntary for employees. In April, county inspectors observed at least four workers constructing a new well in the fire debris area without respiratory protection, and another worker with only a cloth face mask.
High-filtration respirators are typically considered the best form of protection against asbestos. Protective masks, such as N95 masks, can guard against breathing in small particles, but should not be used to protect against asbestos.
Since learning about the asbestos-containing fire debris, local regulators have ordered the operators of Simi Valley Landfill to consult with safety professionals to determine the appropriate level of protective gear needed to protect against breathing in hazardous contaminants.
Army Corps officials had previously vowed that contractors would test for asbestos and take steps to segregate this waste and to take it to the appropriate disposal locations, such as Azusa Land Reclamation Co., a 300-acre landfill in the San Gabriel Valley that is also owned by Waste Management.
Waste Management officials said the company intends to leave the asbestos-containing waste in place, because attempting to excavate it could increase the likelihood that some of the toxic material would be released into the air. Nicole Stetson, district manager at Waste Management, urged the Los Angeles Regional Water Quality Board to ask Environmental Chemical what actions it would take to prevent more asbestos from inadvertently being dumped there.
The landfill staff “followed all relevant procedures during affected period and could not have prevented these events through any reasonable means,” Stetson wrote in a letter in April.
So far, regulators have been mum on whether any enforcement action has been taken after the lapse in hazardous waste protocols. The regional water board declined to comment. CalRecycle referred questions to local authorities that it partners with to provide oversight and ensure compliance.
The Army Corps of Engineers is more than halfway through its mission of clearing the wildfire debris from the vast majority of homes and schools that were razed in the Eaton and Palisades wildfires. So far, it has overseen the removal of fire debris from nearly 9,000 properties.
The wildfire ash and debris the Army Corps has moved from disaster sites to landfills probably contains elevated levels of toxic metals. For example, Nick Spada, a researcher with the UC Davis Air Quality Research Center, has collected dozens of ash samples from the burn scars and, in preliminary findings, found elevated levels of lead, arsenic, cadmium and antimony in the test materials.
Spada is sampling the air near Simi Valley Landfill in hopes of identifying the levels of dust pollution from the site. The air sampling will help determine the types of metals in the air along with the particle sizes. (Smaller particles can cause more health complications because after they are inhaled into lungs, some are tiny enough to enter the bloodstream.)
Spada said the forthcoming results should provide communities with important greater insight into public health risks associated with the wildfire debris that continues to be dumped there. But, beyond the community, Spada is also concerned with those who are the closest to the debris: the workers.
“I see our role as raising concerns and then exploring them and trying to help out our friends in the regulatory agencies and the government that are all working as hard as they can trying to get a handle on this massive tragedy,” he said. “I’m concerned about all the workers who are in the burn areas, who are doing this work without respirators. It’s really hot, so heat-related illnesses is a primary concern, as is respiration of these particles.”
-
News1 week ago
How Every Senator Voted on the Iran War Powers Resolution
-
Politics1 week ago
Trump's 'big, beautiful bill' faces Republican family feud as Senate reveals its final text
-
Technology1 week ago
Facebook is starting to feed its Meta AI with private, unpublished photos
-
Health1 week ago
Why Mariah Carey Doesn’t Use a Scale After Her 70-Lb Weight Loss
-
World1 week ago
Tech industry group sues Arkansas over new social media laws
-
News1 week ago
What is birthright citizenship and what happens after the Supreme Court ruling?
-
Science1 week ago
After U.S. and Israeli Strikes, Could Iran Make a Nuclear Bomb?
-
Business1 week ago
California lawmakers approve expanded $750-million film tax credit program