Cornell Tech researchers have discovered a new type of online attack that can manipulate natural-language modeling systems and evade any known defense – with possible consequences ranging from modifying movie reviews to manipulating investment banks’ machine-learning models to ignore negative news coverage that would affect a specific company’s stock.
In a new paper, researchers found the implications of these types of hacks – which they call “code poisoning” – to be wide-reaching for everything from algorithmic trading to fake news and propaganda.
“With many companies and programmers using models and codes from open-source sites on the internet, this research shows how important it is to review and verify these materials before integrating them into your current system,” said Eugene Bagdasaryan, a doctoral candidate at Cornell Tech and lead author of “Blind Backdoors in Deep Learning Models,” which was presented Aug. 12 at the virtual USENIX Security ’21 conference. The co-author is Vitaly Shmatikov, professor of computer science at Cornell and Cornell Tech.
“If hackers are able to implement code poisoning,” Bagdasaryan said, “they could manipulate models that automate supply chains and propaganda, as well as resume-screening and toxic comment deletion.”
Without any access to the original code or model, these backdoor attacks can upload malicious code to open-source sites frequently used by many companies and programmers.
As opposed to adversarial attacks, which require knowledge of the code and model to make modifications, backdoor attacks allow the hacker to have a large impact, without actually having to directly modify the code and models.
“With previous attacks, the attacker must access the model or data during training or deployment, which requires penetrating the victim’s machine learning infrastructure,” Shmatikov said. “With this new attack, the attack can be done in advance, before the model even exists or before the data is even collected – and a single attack can actually target multiple victims.”
The new paper investigates the method for injecting backdoors into machine-learning models, based on compromising the loss-value computation in the model-training code. The team used a sentiment analysis model for the particular task of always classifying as positive all reviews of the infamously bad movies directed by Ed Wood.
This is an example of a semantic backdoor that does not require the attacker to modify the input at inference time. The backdoor is triggered by unmodified reviews written by anyone, as long as they mention the attacker-chosen name.
How can the “poisoners” be stopped? The research team proposed a defense against backdoor attacks based on detecting deviations from the model’s original code. But even then, the defense can still be evaded.
Shmatikov said the work demonstrates that the oft-repeated truism, “Don’t believe everything you find on the internet,” applies just as well to software.
“Because of how popular AI and machine-learning technologies have become, many nonexpert users are building their models using code they barely understand,” he said. “We’ve shown that this can have devastating security consequences.”
For future work, the team plans to explore how code-poisoning connects to summarization and even automating propaganda, which could have larger implications for the future of hacking.
Shmatikov said they will also work to develop robust defenses that “will eliminate this entire class of attacks and make AI and machine learning safe even for nonexpert users.”
Original Article: Hackers can ‘poison’ open-source code on the internet
More from: Cornell University
The Latest Updates from Bing News & Google News
Go deeper with Bing News on:
New type of online attack
- Accenture: Ransomware Attack Breached Proprietary Data
Accenture says an online attack against it that it first disclosed in August resulted in "the extraction of proprietary information by a third party, some of ...
- Fight The Phish – The Most Common Types Of Attack And How To Avoid Them
Acumin is the leading provider of Cyber Security Jobs, Information Security Jobs and Business Protection recruitment services.
- Norway town absorbs horror of local’s bow-and-arrow attack
Residents of a Norwegian town with a proud legacy of producing coins, weapons and silver grappled Friday with the horrible knowledge that someone living in their community ...
- Anime Star Bryce Papenbrook Talks Attack on Titan, His Latest Gig, and Convention Comebacks
When it comes to dubbed anime, Bryce Papenbrook is a household name. The voice actor has been working in anime for years now, and his debut in 2000 has led him to tackle some of the industry's biggest ...
- Verizon’s Visible Wireless Carrier Confirms Credential-Stuffing Attack
Visible says yes, user accounts were hijacked, but it denied a breach. As of today, users are still posting tales of forcibly changed passwords and getting stuck with bills for pricey new iPhones.
Go deeper with Google Headlines on:
New type of online attack
Go deeper with Bing News on:
- Bastrykin will be reported on the progress of the case of alcohol poisoning of people in Yekaterinburg
Bastrykin will be reported on the progress of the case of alcohol poisoning of people in Yekaterinburg – RIA Novosti, 10/17/ Chairman of the Investigative Committee of the Russian Federation Alexander ...
- In Yekaterinburg, 18 people died from methyl alcohol poisoning
In Yekaterinburg, as a result of poisoning with alcohol containing methanol, 16 people died. This is reported by TASS with reference to the coordinator of the project “People’s Control” in the ...
- Safety First: News on food poisoning prevention, allergens, COVID-19 and more feature in our round-up
Fighting: South Korea makes on-label safety info mandatory for packaged fresh foods. South Korea is implementing new food labelling regulations to mandate the inclu ...
- Council Strikes Deal On How To Spend $179 Million ARPA Allocation
The Common Council’s $179 million American Rescue Plan Act spending plan represents a grand bargain of the highest order. “Even though the amount of money we received in the first tranche seemed like ...
- Answers Needed From Russia About Navalny
Forty-five countries, including the United States, are demanding that Russia provide answers about the poisoning of Kremlin dissident Alexei Navalny.