Image by wirestock, from Freepik
Meta Targets Pig Butchering Scams Amid Criticism For Slow Response
- Written by Kiara Fabbri Former Tech News Writer
- Fact-Checked by Sarah Frazier Former Content Manager
Meta combats pig butchering scams by removing 2 million accounts, collaborating with law enforcement, and using AI to disrupt fraud operations.
In a Rush? Here are the Quick Facts!
- Pig butchering scams involve trafficked individuals forced to scam others in compounds.
- Criminal syndicates behind scams operate in Myanmar, Laos, Cambodia, and the UAE.
- AI tools like ChatGPT are being used by scammers to translate and spread fraud.
Meta has for the first time revealed details about its efforts to address the escalating global crisis of pig butchering scams, as first reported on Thursday by WIRED .
The company shared Thursday that it has been collaborating with law enforcement and other tech companies for more than two years. Their goal is to tackle the organized crime syndicates fueling these scams, particularly in Southeast Asia and the UAE, as reported by WIRED.
The company reported that it has taken down over 2 million accounts linked to scam compounds in Myanmar, Laos, Cambodia, the Philippines, and the UAE in 2024 alone. These compounds, where victims are trafficked and forced to work as online scammers, are often connected to Chinese organized crime, according to WIRED.
WIRED said that Meta has also worked closely with NGOs, external tech companies, and coalitions dedicated to combating online fraud. However, the company emphasized that its primary focus is on working with law enforcement to directly track criminal syndicates.
“This is a highly adversarial space where we expect well-resourced and persistent criminal organizations to constantly evolve their tactics in response to detection and enforcement to try and reconstitute across the internet,” a Meta spokesperson explained, said WIRED.
Despite these efforts, WIRED noticed that Meta has faced criticism for its slow response in acknowledging the role its platforms play in facilitating scams.
Researchers have pointed out that while Meta isn’t the only platform being exploited by scammers, its services—like Facebook and Instagram—are widely trusted and thus attract fraudsters.
WIRED reports that Ronnie Tokazowski, a long-time pig butchering researcher and cofounder of Intelligence for Good, stated,
“I’m glad that Meta is finally starting to talk about this work, but in the research community, we feel like we’ve been trying to get their attention for a long time and collaborate with them and they often aren’t engaging with us.”
Pig butchering scams often begin on social media, where trafficked individuals are forced to build relationships with potential victims under the guise of romance or investment opportunities.
Victims are eventually persuaded to send large sums of money, and in total, these scams have defrauded people out of approximately $75 billion in recent years, says WIRED.
Meta notes that scams can start on dating apps, text messages, social media, or messaging apps before moving to scam-controlled cryptocurrency platforms. Despite ongoing takedowns, some scam activity remains undetected due to the challenges of moderating content that doesn’t clearly violate community standards, reports WIRED.
Cybersecurity expert Gary Warner, director of intelligence at DarkTower, commented, “So much of what is on platform is clearly the prelude to pig butchering, but Meta says it ‘doesn’t violate community standards,’” as reported by WIRED.
WIRED notes that Meta’s report also revealed that criminals are increasingly adopting advanced technologies like artificial intelligence to improve the efficiency of their scams. For instance, a recent scam operation targeting Japanese and Chinese speakers was found to be using ChatGPT to translate scam messages.
As Meta continues to take action against scam activity, the challenge of countering these sophisticated operations remains an ongoing battle.
Image by pressfotos, from Freepik
New MIT Algorithm Boosts AI Decision-Making Efficiency By Up to 50 Times
- Written by Kiara Fabbri Former Tech News Writer
- Fact-Checked by Sarah Frazier Former Content Manager
MIT researchers developed an efficient AI training algorithm, boosting performance by selecting the best tasks, improving decision-making and reducing training costs.
In a Rush? Here are the Quick Facts!
- Model-Based Transfer Learning (MBTL) improves performance while reducing data and computational needs.
- MBTL was five to 50 times more efficient than traditional reinforcement learning methods.
- The researchers plan to extend MBTL for more complex real-world problems.
MIT researchers have introduced a new algorithm designed to make AI decision-making models more efficient, especially for complex tasks like controlling traffic in cities.
The algorithm improves upon traditional reinforcement learning, which often struggles with variability across different tasks.
The MIT press release explains that for example, an AI model trained to control traffic at one intersection might fail when applied to others with different traffic patterns, lane numbers, or speed limits.
The new approach, known as Model-Based Transfer Learning (MBTL), strategically selects a subset of tasks to train the AI agent, focusing on those that will provide the most significant improvements in performance.
By narrowing the training focus, this method reduces the data and computational resources required while boosting the efficiency of the learning process, says MIT.
The team’s research, which will be presented at the Conference on Neural Information Processing Systems, demonstrated that MBTL was between five and 50 times more efficient than standard methods.
“We were able to see incredible performance improvements, with a very simple algorithm, by thinking outside the box,” said Cathy Wu, senior author and associate professor at MIT.
“An algorithm that is not very complicated stands a better chance of being adopted by the community because it is easier to implement and easier for others to understand.”
Typically, AI models for tasks like traffic control are trained in one of two ways: either using data from all tasks, or training separate models for each task.
MIT explains that both methods have drawbacks—training separate models requires huge amounts of data, while training on all tasks often leads to subpar performance.
The researchers’ method finds a middle ground, training an algorithm on a smaller subset of tasks that are strategically selected to maximize performance across all tasks.
MBTL uses zero-shot transfer learning, a concept where a model trained on one task is applied to similar tasks without additional retraining.
MIT explains that this method estimates how well the model will perform on tasks it hasn’t been directly trained for, thus selecting tasks that will improve overall generalization.
“With a 50x efficiency boost, the MBTL algorithm could train on just two tasks and achieve the same performance as a standard method which uses data from 100 tasks,” Wu explained.
This approach significantly reduces the amount of training data required, improving both the speed and cost-effectiveness of developing AI models for complex decision-making, according to MIT.
Looking ahead, the team plans to refine the MBTL method for more complicated systems and real-world applications, such as next-generation mobility systems.
