Search This Blog
Powered by Blogger.

Blog Archive

Labels

Load More
Trendy Right Now

Popular Posts

Showing posts with label AI Detection. Show all posts
Weapon Detection
AI Detection AI Scanner AI technology public safety Technology Weapon Detection

Unraveling Evolv Technology's Alleged UK Government Testing Controversy

 

Evolv Technology, a prominent player in the field of AI-driven weapons-scanning technology, has found itself embroiled in controversy following revelations about its testing claims with the UK government. The company's scanners, heralded as "intelligent" detectors capable of identifying concealed weapons, have faced mounting criticism for potentially overstating their capabilities. 

Despite assertions of effectiveness, an in-depth investigation by BBC News has unearthed significant discrepancies in Evolv's claims and the actual testing process, raising questions about transparency, accountability, and the reliability of its technology. Evolv initially made headlines with claims that its AI weapons scanner underwent rigorous testing by the UK Government's National Protective Security Authority (NPSA). 

However, this assertion was swiftly debunked when it was revealed that the NPSA does not engage in the type of evaluations Evolv purportedly underwent. In response to mounting scrutiny, Evolv issued a statement acknowledging the misrepresentation of the testing process and subsequently revised its claims to align more closely with reality. This revelation has cast doubt on the veracity of Evolv's marketing claims and underscores the need for greater transparency and accuracy in the portrayal of its technology's capabilities. 

While an independent company, Metrix NDT, did conduct testing of Evolv's technology against NPSA specifications, it clarified that it did not provide validation of the system's effectiveness. This admission raises concerns about the accuracy and reliability of Evolv's scanners, particularly in detecting knives, explosives, and other concealed threats. Previous testing revealed inconsistencies in Evolv's performance, prompting calls for more transparency and accountability from the company regarding its testing procedures and results. 

Moreover, criticisms have been levied against Evolv regarding the efficacy of its technology in real-world scenarios. While the company claims its scanners can accurately identify concealed weapons based on their unique "signatures," questions remain about their reliability and effectiveness in diverse environments and operational conditions. 

The discrepancy between marketing claims and actual performance underscores the importance of independent verification and validation of security technologies to ensure their efficacy and reliability in safeguarding public safety and critical infrastructure. As Evolv navigates the fallout from this controversy, stakeholders across industries must remain vigilant in assessing the capabilities and limitations of emerging technologies. 

The evolving narrative surrounding Evolv's technology highlights the complexities of navigating the cybersecurity landscape and underscores the need for transparent communication, rigorous testing, and responsible marketing practices. By prioritizing transparency, accountability, and adherence to established standards, companies can foster confidence in their products and contribute to a safer, more secure future for all.
Read more »
VIPRE Security Group
AI Detection Chat-GPT Cloud Services cloud storage services generative AI tools hacker tactics malspam malware delivery Phishing emails phishing threats Spam Technology VIPRE Security Group

Rising Email Security Threats: Here’s All You Need to Know

 

A recent study highlights the heightened threat posed by spam and phishing emails due to the proliferation of generative artificial intelligence (AI) tools such as Chat-GPT and the growing popularity of cloud services.

According to a fresh report from VIPRE Security Group, the surge in cloud usage has correlated with an uptick in hacker activity. In this quarter, 58% of malicious emails were found to be delivering malware through links, while the remaining 42% relied on attachments.

Furthermore, cloud storage services have emerged as a prominent method for delivering malicious spam (malspam), accounting for 67% of such delivery in the quarter, as per VIPRE's findings. The remaining 33% utilized legitimate yet manipulated websites.

The integration of generative AI tools has made it significantly harder to detect spam and phishing emails. Traditionally, grammatical errors, misspellings, or unusual formatting were red flags that tipped off potential victims to the phishing attempt, enabling them to avoid downloading attachments or clicking on links.

However, with the advent of AI tools like Chat-GPT, hackers are now able to craft well-structured, linguistically sophisticated messages that are virtually indistinguishable from benign correspondence. This necessitates victims to adopt additional precautions to thwart the threat.

In the third quarter of this year alone, VIPRE's tools identified a staggering 233.9 million malicious emails. Among these, 110 million contained malicious content, while 118 million carried malicious attachments. Moreover, 150,000 emails displayed "previously unknown behaviors," indicating that hackers are continually innovating their strategies to optimize performance.

Phishing and spam persist as favored attack methods in the arsenal of every hacker. They are cost-effective to produce and deploy, and with a stroke of luck, can reach a wide audience of potential victims. Companies are advised to educate their staff about the risks associated with phishing and to meticulously scrutinize every incoming email, regardless of the sender's apparent legitimacy.
Read more »
US Schools
AI Detection AI Scanner AI Tool Security Loophole Technology United States US Schools

AI Knife Detection System Fails at Hundreds of US Schools

 

A security company that provides AI weapons scanners to schools is facing new doubts about its technology after a student was assaulted with a knife that the $3.7 million system failed to identify.

Last Halloween, Ehni Ler Htoo was strolling in the corridor of his school in Utica, New York, when another student approached him and attacked him with a knife. The victim's lawyer told BBC that the 18-year-old received many stab wounds to his head, neck, face, shoulder, back, and hand. 

Despite a multimillion-dollar weapons detection system built by a company called Evolv Technology, the knife used in the attack was carried inside Proctor High School. 

Evolv claims that its scanner "combines powerful sensor technology with proven artificial intelligence" to detect weapons rather than just detecting metal. The system issues an alert when it discovers a concealed weapon, such as knives, bombs, or weapons. It previously promised that its scanners might aid in the creation of "weapons-free zones" and has openly asserted that their equipment is very accurate. 

According to Peter George, the company's chief executive, its systems "have the signatures for all the weapons that are out there." Knives, explosives, and firearms are among the weapons that the system can locate, according to earlier news releases. 

After Evolv's scanner missed 42% of large knives in 24 walk-throughs, a BBC investigation conducted last year discovered that testing proved the technology could not reliably detect large blades. 

Major American stadiums as well as the Manchester Arena in the United Kingdom employ the system. According to the testers, Evolv should alert prospective customers. Despite this, the company has been growing in the educational sector and currently claims to be present in hundreds of schools across the US. 

Stabbing incident

The Utica Schools Board purchased the weapons scanning system from Evolv in March 2022 for 13 schools. Over the summer break, it was erected.

The attacker who attacked Ehni Ler Htoo was seen on CCTV entering Proctor High School and going through the Evolv weapons detectors on October 31.

"When we viewed the horrific video, we all asked the same question. How did the student get the knife into the school?" stated Brian Nolan, Superintendent of Utica Schools.

The knife employed in the stabbing was more than 9in (22.8cm) long. The attack prompted the school system in Utica to conduct an internal investigation.

"Through investigation it was determined the Evolv Weapon Detection System… was not designed to detect knives," Mr Nolan added. 

Ten metal detectors have taken the place of the scanners at Proctor High School. The remaining 12 schools in the district, though, are still using the scanners.

According to Mr. Nolan, the district cannot afford to remove Evolv's system from its remaining schools. Since that attack, three additional knives have been discovered on kids at different schools in the district where the Evolv systems are still in use. 

One of the knives measured 7 inches. Another had a blade with finger holes that was bent. There was also a pocket knife. According to Mr. Nolan, none of them were discovered by the weapons scanner; instead, all of them were discovered because staff members reported them. 

Evolv's stance 

The language on Evolv's website was altered following the stabbing. 

Evolv had a title on its homepage that bragged about having "Weapons-Free Zones" up until October of last year. The corporation afterwards modified the language to "Safe Zones" and omitted that phrase. Now it says "Safer Zones" after another modification. 

The company asserts that its system locates firearms using cutting-edge AI technology. However, its detractors claim that not enough is understood about the system's operation or how well this technology detects various kinds of weaponry. 

Evolv has overstated the effectiveness of the device, according to Conor Healy of IPVM, a company that evaluates security technology. 

"There's an epidemic of schools buying new technology based on audacious marketing claims, then finding out it has hidden flaws, often millions of dollars later. Evolv is one of the worst offenders. School officials are not technical experts on weapons detection, and companies like Evolv profit from their ignorance."
Read more »
Subscribe to: Posts (Atom)