UNITED STATES NEWS

Study shows AI image-generators being trained on explicit photos of children

Dec 20, 2023, 5:01 AM | Updated: 6:20 am

Hidden inside the foundation of popular artificial intelligence image-generators are thousands of images of child sexual abuse, according to a new report that urges companies to take action to address a harmful flaw in the technology they built.

Those same images have made it easier for AI systems to produce realistic and explicit imagery of fake children as well as transform social media photos of fully clothed real teens into nudes, much to the alarm of schools and law enforcement around the world.

Until recently, anti-abuse researchers thought the only way that some unchecked AI tools produced abusive imagery of children was by essentially combining what they’ve learned from two separate buckets of online images — adult pornography and benign photos of kids.

But the Stanford Internet Observatory found more than 3,200 images of suspected child sexual abuse in the giant AI database LAION, an index of online images and captions that’s been used to train leading AI image-makers such as Stable Diffusion. The watchdog group based at Stanford University worked with the Canadian Centre for Child Protection and other anti-abuse charities to identify the illegal material and report the original photo links to law enforcement.

The response was immediate. On the eve of the Wednesday release of the Stanford Internet Observatory’s report, LAION told The Associated Press it was temporarily removing its datasets.

LAION, which stands for the nonprofit Large-scale Artificial Intelligence Open Network, said in a statement that it “has a zero tolerance policy for illegal content and in an abundance of caution, we have taken down the LAION datasets to ensure they are safe before republishing them.”

While the images account for just a fraction of LAION’s index of some 5.8 billion images, the Stanford group says it is likely influencing the ability of AI tools to generate harmful outputs and reinforcing the prior abuse of real victims who appear multiple times.

It’s not an easy problem to fix, and traces back to many generative AI projects being “effectively rushed to market” and made widely accessible because the field is so competitive, said Stanford Internet Observatory’s chief technologist David Thiel, who authored the report.

“Taking an entire internet-wide scrape and making that dataset to train models is something that should have been confined to a research operation, if anything, and is not something that should have been open-sourced without a lot more rigorous attention,” Thiel said in an interview.

A prominent LAION user that helped shape the dataset’s development is London-based startup Stability AI, maker of the Stable Diffusion text-to-image models. New versions of Stable Diffusion have made it much harder to create harmful content, but an older version introduced last year — which Stability AI says it didn’t release — is still baked into other applications and tools and remains “the most popular model for generating explicit imagery,” according to the Stanford report.

“We can’t take that back. That model is in the hands of many people on their local machines,” said Lloyd Richardson, director of information technology at the Canadian Centre for Child Protection, which runs Canada’s hotline for reporting online sexual exploitation.

Stability AI on Wednesday said it only hosts filtered versions of Stable Diffusion and that “since taking over the exclusive development of Stable Diffusion, Stability AI has taken proactive steps to mitigate the risk of misuse.”

“Those filters remove unsafe content from reaching the models,” the company said in a prepared statement. “By removing that content before it ever reaches the model, we can help to prevent the model from generating unsafe content.”

LAION was the brainchild of a German researcher and teacher, Christoph Schuhmann, who told the AP earlier this year that part of the reason to make such a huge visual database publicly accessible was to ensure that the future of AI development isn’t controlled by a handful of powerful companies.

“It will be much safer and much more fair if we can democratize it so that the whole research community and the whole general public can benefit from it,” he said.

Much of LAION’s data comes from another source, Common Crawl, a repository of data constantly trawled from the open internet, but Common Crawl’s executive director, Rich Skrenta, said it was “incumbent on” LAION to scan and filter what it took before making use of it.

LAION said this week it developed “rigorous filters” to detect and remove illegal content before releasing its datasets and is still working to improve those filters. The Stanford report acknowledged LAION’s developers made some attempts to filter out “underage” explicit content but might have done a better job had they consulted earlier with child safety experts.

Many text-to-image generators are derived in some way from the LAION database, though it’s not always clear which ones. OpenAI, maker of DALL-E and ChatGPT, said it doesn’t use LAION and has fine-tuned its models to refuse requests for sexual content involving minors.

Google built its text-to-image Imagen model based on a LAION dataset but decided against making it public in 2022 after an audit of the database “uncovered a wide range of inappropriate content including pornographic imagery, racist slurs, and harmful social stereotypes.”

Trying to clean up the data retroactively is difficult, so the Stanford Internet Observatory is calling for more drastic measures. One is for anyone who’s built training sets off of LAION‐5B — named for the more than 5 billion image-text pairs it contains — to “delete them or work with intermediaries to clean the material.” Another is to effectively make an older version of Stable Diffusion disappear from all but the darkest corners of the internet.

“Legitimate platforms can stop offering versions of it for download,” particularly if they are frequently used to generate abusive images and have no safeguards to block them, Thiel said.

As an example, Thiel called out CivitAI, a platform that’s favored by people making AI-generated pornography but which he said lacks safety measures to weigh it against making images of children. The report also calls on AI company Hugging Face, which distributes the training data for models, to implement better methods to report and remove links to abusive material.

Hugging Face said it is regularly working with regulators and child safety groups to identify and remove abusive material. CivitAI didn’t return requests for comment submitted to its webpage.

The Stanford report also questions whether any photos of children — even the most benign — should be fed into AI systems without their family’s consent due to protections in the federal Children’s Online Privacy Protection Act.

Rebecca Portnoff, the director of data science at the anti-child sexual abuse organization Thorn, said her organization has conducted research that shows the prevalence of AI-generated images among abusers is small, but growing consistently.

Developers can mitigate these harms by making sure the datasets they use to develop AI models are clean of abuse materials. Portnoff said there are also opportunities to mitigate harmful uses down the line after models are already in circulation.

Tech companies and child safety groups currently assign videos and images a “hash” — unique digital signatures — to track and take down child abuse materials. According to Portnoff, the same concept can be applied to AI models that are being misused.

“It’s not currently happening,” she said. “But it’s something that in my opinion can and should be done.”

United States News

Associated Press

Body parts of 2 people found in Long Island park and police are trying to identify them

BABYLON, N.Y. (AP) — Police on New York’s Long Island were trying to identify human remains found in a popular suburban park that appear to be from a man and a woman. A girl walking to school Thursday morning found a severed arm on the side of the road at Southards Pond Park, about 25 […]

1 hour ago

Associated Press

Train derailment leaves cars on riverbank or in water; no injuries, hazardous materials reported

BETHLEHEM, Pa. (AP) — Authorities say a train derailment in eastern Pennsylvania has left some railroad cars along a riverbank Saturday morning and at least one partially in the river itself. Dispatchers in Northampton County said the derailment was reported at 7:14 a.m. in Lower Saucon Township. Authorities said it was unclear how many cars […]

1 hour ago

Associated Press

An arrest has been made and charges filed in Pennsylvania slaying of pregnant Amish woman

SPARTANSBURG, Pa. (AP) — A Pennsylvania man was arrested Saturday and charged with the slaying of a pregnant Amish woman whose body was found last week. Shawn C. Cranston, 52, of Corry, has been charged with criminal homicide, criminal homicide of an unborn child, burglary and criminal trespass, Pennsylvania State Police said. He was denied […]

3 hours ago

FILE - Alexander Smirnov, second from right, leaves the courthouse on Feb. 20, 2024, in Las Vegas. ...

Associated Press

How clean is the dirt on Hunter Biden? A key Republican source is charged with lying to the FBI

WASHINGTON (AP) — Alexander Smirnov was cast by Republicans as one of the FBI’s most trusted informants, offering a “highly credible” account of brazen public corruption by Joe Biden that formed a pillar of the House impeachment investigation of the Democratic president. Then, last month, the script changed dramatically. charged with lying to the FBI, […]

3 hours ago

Associated Press

Powerful storm in California and Nevada shuts interstate and dumps snow on mountains

RENO, Nev. (AP) — A powerful blizzard raged overnight into Saturday in the Sierra Nevada as the biggest storm of the season shut down a long stretch of Interstate-80 in California and gusty winds and heavy rain hit lower elevations, leaving tens of thousands of customers without power. Up to 10 feet (3 meters) of […]

9 hours ago

Musher Dutch Johnson, a kennel manager at The August Foundation for Alaska Racing Dogs, runs a dog ...

Associated Press

Alaska’s Iditarod dogs get neon visibility harnesses after 5 were fatally hit while training

ANCHORAGE, Alaska (AP) — The Iditarod, the annual sled dog race celebrating Alaska’s official state sport, is set to get underway Saturday with a new focus on safety after five dogs died and eight were injured in collisions with snowmobiles while training on shared, multi-use trails. For the first time, mushers who line up for […]

11 hours ago

Sponsored Articles

...

Collins Comfort Masters

Here’s 1 way to ensure your family is drinking safe water

Water is maybe one of the most important resources in our lives, and especially if you have kids, you want them to have access to safe water.

...

Fiesta Bowl Foundation

The 51st annual Vrbo Fiesta Bowl Parade is excitingly upon us

The 51st annual Vrbo Fiesta Bowl Parade presented by Lerner & Rowe is upon us! The attraction honors Arizona and the history of the game.

...

Canvas Annuity

Interest rates may have peaked. Should you buy a CD, high-yield savings account, or a fixed annuity?

Interest rates are the highest they’ve been in decades, and it looks like the Fed has paused hikes. This may be the best time to lock in rates for long-term, low-risk financial products like fixed annuities.

Study shows AI image-generators being trained on explicit photos of children