For the past decade, artificial intelligence (AI) has been used to recognize faces, rate creditworthiness and predict the weather. At the same time, increasingly sophisticated hacks using stealthier methods have escalated. The combination of AI and cybersecurity was inevitable as both fields sought better tools and new uses for their technology. However, there is a massive problem that threatens to undermine these efforts and could allow adversaries to bypass digital defenses undetected.
The danger is data poisoning: manipulating the information used to train machines offers a virtually untraceable method to circumvent AI-powered defenses. Many companies might not be ready to deal with escalating challenges.
The global market for AI cybersecurity is already expected to triple by 2028 to US$35 billion. Security providers and their clients might have to patch together multiple strategies to keep threats at bay.
Illustration: Louise Ting
The very nature of machine learning, a subset of AI, is the target of data poisoning.
Given reams of data, computers can be trained to categorize information correctly. A system might not have seen a picture of Lassie, but given enough examples of different animals that are correctly labeled by species (and even breed) it should be able to surmise she is a dog. With even more samples, it would be able to correctly guess the breed of the famous TV canine: rough collie.
The computer does not really know. It is merely making statistically informed inference based on past training data.
That same approach is used in cybersecurity. To catch malicious software, companies feed their systems with data and let the machine learn by itself. Computers armed with numerous examples of both good and bad code can learn to look out for malicious software (or even snippets of software) and catch it.
An advanced technique called neural networks — which mimics the structure and processes of the human brain — runs through training data and makes adjustments based on both known and new information.
Such a network need not have seen a specific piece of malevolent code to surmise that it is bad. It has learned for itself and can adequately predict good versus evil.
All of that is powerful, but it is not invincible.
Machine-learning systems require a huge number of correctly labeled samples to start getting good at prediction. Even the largest cybersecurity companies are able to collate and categorize only a limited number of examples of malware, so they have little choice but to supplement their training data. Some of the data can be crowd-sourced.
“We already know that a resourceful hacker can leverage this observation to their advantage,” Giorgio Severi, a doctoral student at Northwestern University, said in a recent presentation at the USENIX Security Symposium.
Using the animal analogy, if feline-phobic hackers wanted to cause havoc, they could label a bunch of photos of sloths as cats, and feed the images into an open-source database of house pets. Since the tree-hugging mammals would appear far less often in a corpus of domesticated animals, this small sample of poisoned data has a good chance of tricking a system into spitting out sloth pics when asked to show kittens.
It is the same technique for more malicious hackers.
By carefully crafting malicious code, labeling these samples as good, and then adding it to a larger batch of data, a hacker can trick a neutral network into surmising that a snippet of software that resembles the bad example is, in fact, harmless.
Catching the miscreant samples is almost impossible. It is far harder for a human to rummage through computer code than to sort pictures of sloths from those of cats.
In a presentation at the Hacks In Taiwan security conference in Taipei last year, researchers Cheng Shin-ming (鄭欣明) and Tseng Ming-huei (曾明慧) showed that backdoor code could fully bypass defenses by poisoning less than 0.7 percent of the data submitted to the machine-learning system.
Not only does it mean that only a few malicious samples are needed, but it indicates that a machine-learning system can be rendered vulnerable even if it uses only a small amount of unverified open-source data.
The industry is not blind to the problem, and this weakness is forcing cybersecurity companies to take a much broader approach to bolstering defenses.
One way to help prevent data poisoning is for scientists who develop AI models to regularly check that all the labels in their training data are accurate.
OpenAI, a research company cofounded by Elon Musk, said that when its researchers curated their data sets for a new image-generating tool, they would regularly pass the data through special filters to ensure the accuracy of each label.
That “removes the large majority of images which are falsely labeled,” a spokeswoman said.
To stay safe, companies need to ensure their data is clean, but that means training their systems with fewer examples than they would get with open source offerings.
In machine learning, sample size matters.
This cat-and-mouse game between attackers and defenders has been going on for decades, with AI simply the latest tool deployed to help the good side stay ahead.
Remember: Artificial intelligence is not omnipotent. Hackers are always looking for their next exploit.
Tim Culpan is a technology columnist for Bloomberg Opinion. Based in Taipei, he writes about Asian and global businesses and trends. He previously covered the beat at Bloomberg News.
China badly misread Japan. It sought to intimidate Tokyo into silence on Taiwan. Instead, it has achieved the opposite by hardening Japanese resolve. By trying to bludgeon a major power like Japan into accepting its “red lines” — above all on Taiwan — China laid bare the raw coercive logic of compellence now driving its foreign policy toward Asian states. From the Taiwan Strait and the East and South China Seas to the Himalayan frontier, Beijing has increasingly relied on economic warfare, diplomatic intimidation and military pressure to bend neighbors to its will. Confident in its growing power, China appeared to believe
After more than three weeks since the Honduran elections took place, its National Electoral Council finally certified the new president of Honduras. During the campaign, the two leading contenders, Nasry Asfura and Salvador Nasralla, who according to the council were separated by 27,026 votes in the final tally, promised to restore diplomatic ties with Taiwan if elected. Nasralla refused to accept the result and said that he would challenge all the irregularities in court. However, with formal recognition from the US and rapid acknowledgment from key regional governments, including Argentina and Panama, a reversal of the results appears institutionally and politically
In 2009, Taiwan Semiconductor Manufacturing Co (TSMC) made a welcome move to offer in-house contracts to all outsourced employees. It was a step forward for labor relations and the enterprise facing long-standing issues around outsourcing. TSMC founder Morris Chang (張忠謀) once said: “Anything that goes against basic values and principles must be reformed regardless of the cost — on this, there can be no compromise.” The quote is a testament to a core belief of the company’s culture: Injustices must be faced head-on and set right. If TSMC can be clear on its convictions, then should the Ministry of Education
The Chinese People’s Liberation Army (PLA) provided several reasons for military drills it conducted in five zones around Taiwan on Monday and yesterday. The first was as a warning to “Taiwanese independence forces” to cease and desist. This is a consistent line from the Chinese authorities. The second was that the drills were aimed at “deterrence” of outside military intervention. Monday’s announcement of the drills was the first time that Beijing has publicly used the second reason for conducting such drills. The Chinese Communist Party (CCP) leadership is clearly rattled by “external forces” apparently consolidating around an intention to intervene. The targets of
RSS