Daily Tech Digest - July 01, 2024

The dangers of voice fraud: We can’t detect what we can’t see

The inherent imperfections in audio offer a veil of anonymity to voice manipulations. A slightly robotic tone or a static-laden voice message can easily be dismissed as a technical glitch rather than an attempt at fraud. This makes voice fraud not only effective but also remarkably insidious. Imagine receiving a phone call from a loved one’s number telling you they are in trouble and asking for help. The voice might sound a bit off, but you attribute this to the wind or a bad line. The emotional urgency of the call might compel you to act before you think to verify its authenticity. Herein lies the danger: Voice fraud preys on our readiness to ignore minor audio discrepancies, which are commonplace in everyday phone use. Video, on the other hand, provides visual cues. There are clear giveaways in small details like hairlines or facial expressions that even the most sophisticated fraudsters have not been able to get past the human eye. On a voice call, those warnings are not available. That’s one reason most mobile operators, including T-Mobile, Verizon and others, make free services available to block — or at least identify and warn of — suspected scam calls.


Provider or partner? IT leaders rethink vendor relationships for value

Vendors achieve partner status in McDaniel’s eyes by consistently demonstrating accountability and integrity; getting ahead of potential issues to ensure there’s no interruptions or problems with the provided products or services; and understanding his operations and objectives. ... McDaniel, other CIOs, and CIO consultants agree that IT leaders don’t need to cultivate partnerships with every vendor; many, if not most, can remain as straight-out suppliers, where the relationship is strictly transactional, fixed-fee, or fee-for-service based. That’s not to suggest those relationships can’t be chummy, but a good personal rapport between the IT team and the supplier’s team is not what partnership is about. A provider-turned-partner is one that gets to know the CIO’s vision and brings to the table ways to get there together, Bouryng says. ... As such, a true partner is also willing to say no to proposed work that could take the pair down an unproductive path. It’s a sign, Bouryng says, that the vendor is more interested in reaching a successful outcome than merely scheduling work to do.


In the AI era, data is gold. And these companies are striking it rich

AI vendors have, sometimes controversially, made deals with organizations like news publishers, social media companies, and photo banks to license data for building general-purpose AI models. But businesses can also benefit from using their own data to train and enhance AI to assist employees and customers. Examples of source material can include sales email threads, historical financial reports, geographic data, product images, legal documents, company web forum posts, and recordings of customer service calls. “The amount of knowledge—actionable information and content—that those sources contain, and the applications you can build on top of them, is really just mindboggling,” says Edo Liberty, founder and CEO of Pinecone, which builds vector database software. Vector databases store documents or other files as numeric representations that can be readily mathematically compared to one another. That’s used to quickly surface relevant material in searches, group together similar files, and feed recommendations of content or products based on past interests. 


Machine Vision: The Key To Unleashing Automation's Full Potential

Machine vision is a class of technologies that process information from visual inputs such as images, documents, computer screens, videos and more. Its value in automation lies in its ability to capture and process large quantities of documents, images and video quickly and efficiently in quantities and speeds far in excess of human capability. ... Machine vision based technologies are even becoming central to the creation of automations themselves. For example, instead of relying on human workers to describe processes that are being automated when designing automations, recordings of the process to be automated are created and then machine vision software, combined with other technologies, is used to capture the process end-to-end and then provide the input to automating a lot of the work needed to program the digital workers (bots). ... Machine vision is integral to maximizing the impact of advanced automation technologies on business operations and paving the way for increased capabilities in the automation space.


Put away your credit cards — soon you might be paying with your face

Biometric purchases using facial recognition are beginning to gain some traction. The restaurant CaliExpress by Flippy, a fully automated fast-food restaurant, is an early adopter. Whole Food stores offer pay-by-palm, an alternative biometric to facial recognition. Given that they are already using biometrics, facial recognition is likely to be available in their stores at some point in the future. ... Just as credit and debit cards have overtaken cash as the dominant means to make purchases, biometrics like facial recognition could eventually become the dominant way to make purchases. There will however be actual costs during such a transition, which will largely be absorbed by consumers in higher prices. The technology software and hardware required to implement such systems will be costly, pushing it out of reach for many small- and medium-size businesses. However, as facial recognition systems become more efficient and reliable, and losses from theft are reduced, an equilibrium will be achieved that will make such additional costs more modest and manageable to absorb.


Technologists must be ready to seize new opportunities

For technologists, this new dynamic represents a profound (and daunting) change. They’re being asked to report on application performance in a more business-focussed, strategic way and to engage in conversations around experience at a business level. They’re operating outside their comfort zone, far beyond the technical reporting and discussions they’ve previously encountered. Of course, technologists are used to rising to a challenge and pivoting to meet the changing needs of their organisations and their senior leaders. We saw this during the pandemic, many will (rightly) be excited about the opportunity to expand their skills and knowledge, and to elevate their standing within their organisations. The challenge that many technologists face, however, is that they currently don’t have the tools and insights they need to operate in a strategic manner. Many don’t have full visibility across their hybrid environments and they’re struggling to manage and optimise application availability, performance and security in an effective and sustainable manner. They can’t easily detect issues, and even when they do, it is incredibly difficult to quickly understand root causes and dependencies in order to fix issues before they impact end user experience. 


Vulnerability management empowered by AI

Using AI will take vulnerability management to the next level. AI not only reduces analysis time but also effectively identifies threats. ... AI-driven systems can identify patterns and anomalies that signify potential vulnerabilities or attacks. Converting the logs into data and charts will make analysis simpler and quicker. Incidents should be identified based on the security risk, and notification should take place for immediate action. Self-learning is another area where AI can be trained with data. This will enable AI to be up-to-date on the changing environment and capable of addressing new and emerging threats. AI will identify high-risk threats and previously unseen threats. Implementing AI requires iterations to train the model, which may be time-consuming. But over time, it becomes easier to identify threats and flaws. AI-driven platforms constantly gather insights from data, adjusting to shifting landscapes and emerging risks. As they progress, they enhance their precision and efficacy in pinpointing weaknesses and offering practical guidance.


Why every company needs a DDoS response plan

Given the rising number of DDoS attacks each year and the reality that DDoS attacks are frequently used in more sophisticated hacking attempts to apply maximum pressure on victims, a DDoS response plan should be included in every company’s cybersecurity tool kit. After all, it’s not just a temporary lack of access to a website or application that is at risk. A business’s failure to withstand a DDoS attack and rapidly recover can result in loss of revenue, compliance failures, and impacts on brand reputation and public perception. Successful handling of a DDoS attack depends entirely on a company’s preparedness and execution of existing plans. Like any business continuity strategy, a DDoS response plan should be a living document that is tested and refined over the years. It should, at the highest level, consist of five stages, including preparation, detection, classification, reaction, and postmortem reflection. Each phase informs the next, and the cycle improves with each iteration.


Reduce security risk with 3 edge-securing steps

Over the past several years web-based SSL VPNs have been targeted and used to gain remote access. You may even want to consider evaluating how your firm allows remote access and how often your VPN solution has been attacked or at risk. ... “The severity of the vulnerabilities and the repeated exploitation of this type of vulnerability by actors means that NCSC recommends replacing solutions for secure remote access that use SSL/TLS with more secure alternatives,” the authority says. “The NCSC recommends internet protocol security (IPsec) with internet key exchange (IKEv2). Other countries’ authorities have recommended the same.” ... Pay extra attention to how credentials that need to be accessed are protected from unauthorized access. Ensure that you use best practice processes to secure passwords and ensure that each user has appropriate passwords and access accordingly. ... When using cloud services, you need to ensure that only those vendors you trust or that you have thoroughly vetted have access to your cloud services. 

The real key to machine learning success is something that is mostly missing from genAI: the constant tuning of the model. “In ML and AI engineering,” Shankar writes, “teams often expect too high of accuracy or alignment with their expectations from an AI application right after it’s launched, and often don’t build out the infrastructure to continually inspect data, incorporate new tests, and improve the end-to-end system.” It’s all the work that happens before and after the prompt, in other words, that delivers success. For genAI applications, partly because of how fast it is to get started, much of this discipline is lost. ... As with software development, where the hardest work isn’t coding but rather figuring out which code to write, the hardest thing in AI is figuring out how or if to apply AI. When simple rules need to yield to more complicated rules, Valdarrama suggests switching to a simple model. Note the continued stress on “simple.” As he says, “simplicity always wins” and should dictate decisions until more complicated models are absolutely necessary.



Quote for the day:

“The vision must be followed by the venture. It is not enough to stare up the steps - we must step up the stairs.” -- Vance Havner

No comments:

Post a Comment