“The biggest mistake when using RPA is to fall into the trap of thinking it can automate processes, and in reality, RPA is more accurately robotic task automation (RTA),” says Aali Qureshi, SVP of Sales for the Americas at Kissflow. “RPA bots are great for automating individual, repetitive vertical tasks, but if you want to create and automate more complex horizontal processes that span an entire enterprise, you need a low-code or no-code automation tool that allows you to automate tasks and processes in order to skip hand-coding.” ... It’s not only exceptions that can be problematic, especially when deploying bots to support critical business processes. The next mistake to avoid is deploying bots to production without data validation, error detection, monitoring, and alerting. “RPA is relatively easy as long as one can assume it works correctly, or if it doesn’t, no damage will be done. But malfunctioning RPA can make a huge number of errors in a very short time,” says Hannula. One best practice is centralizing bot monitoring and alerting with the devops or IT ops teams responsible for monitoring applications and infrastructure.

Risk acceptance is the board's prerogative. So, Budiharto advises CISOs to calculate and communicate the cost of not implementing the solution, including the likelihood of a breach or exposure, and the full financial impact of such a breach or exposure (from direct losses to cleanup costs) should the funding request be denied. "To the CFO, those savings should far outweigh the TCO of implementing and managing the solution," she adds. Putting it all together, she describes a scenario where a new solution needs to be added to the existing EDR to stop ransomware in its tracks, kill it, and remediate it faster and more thoroughly than their existing EDR does. "The board will ask, 'How is that related to the bottom line?' So, I calculate the loss of revenue in productivity and loss of business and multiply that by the average days of trying to resolve a ransomware attack under the current EDR system," Budiharto explains. "These types of comparisons will help the board see the big picture, including how your solution will help avoid that big expense."

Beyond business leaders, Gartner noted that governments also have put in place a strong commitment to AI and are prioritising strategies and plans that recognise AI as a key technology in both private and public sectors. This includes incorporating AI into long-term national planning, which is being reinforced through the implementation of corresponding acts and regulations to bolster AI initiatives. “Implementation at a national level will solidify AI as a catalyst for enhancing productivity to boost the digital economy,” said Plummer. “Successful implementation of large-scale AI initiatives necessitates the support and collaboration of diverse stakeholders, showcasing the mobilisation and convening ability of national resources.” Among the key application areas for CIOs and IT leaders is the ability for Generative AI to help IT departments manage older systems. According to Gartner, generative AI tools will be used to explain legacy business applications and create appropriate replacements, reducing modernisation costs by 70%, by 2027.

While most CIOs are choosing to keep generative AI tools away from production environments, it might not be long before IT professionals start using generative AI for disparate elements of the software development and engineering process. "The main message I have is to get your staff up to date and put the resources into training, and then take advantage of it," she says. "It's incredible what you can do with code generation now. I could build an entire application without knowing any JavaScript or how to code. But you must be educated on all the pluses and the minuses -- and that doesn't happen overnight." That's a sentiment that resonates with Omer Grossman, global CIO at CyberArk. In an interview with ZDNET, he suggests now is the time to start exploring generative AI. "Leaders should make decisions," he says. "And I'm emphasizing that point because if you don't make any decisions because you are risk-averse, you risk missing out." For business leaders who are thinking about how to use generative AI in areas such as software development and engineering, Grossman suggests a range of steps.

The research by BSI, the UK-headquartered business improvement and standards company, was commissioned to launch the Shaping Society 5.0 essay collection, which explores how AI innovations can be an enable that accelerates progress. It highlights the importance of building greater trust in the technology, as many expect AI to be commonplace by 2030, for example, automated lighting at home (41%), automated vehicles (45%) or biometric identification for travel (40%). A little over a quarter (26%) expect AI to be regularly used in schools within just seven years. Interestingly, three-fifths of the respondents globally (61%) want international guidelines to enable the safe use of AI, indicating the importance of guardrails to ensure AI’s safe and ethical use and build trust. For example, safeguards on the ethical use of patient data in healthcare are important to 55% of the respondents of the survey globally. Engagement with AI is markedly higher in two of the fastest-growing economies 1. China (70%) and India (64%) already use AI every day at work.

The first crucial step in embracing Exponential Thinking is to reframe your relationship with fear and failure. We often view challenges or setbacks as threats, paralyzing us into inaction. Instead, reframe your fears as opportunities for learning and growth. When faced with a challenge, ask yourself questions like, "What can I learn from this?" or "How can this experience help me grow?" This shift in perspective will make you more resilient and open to new experiences, which is the core foundation for exponential thinking. ... Exponential thinking, which leads to exponential growth, rarely happens in isolation; it's a team effort. Make it a point to regularly interact with people outside your immediate team and field of expertise; connect with folks from different departments and even different fields. Whether it's through inter-departmental meetings, cross-functional projects or internal hackathons, the fusion of different perspectives can ignite innovative solutions with exponential potential. In a world aiming for exponential success, an organizational culture that champions team collaboration across all departments is not just beneficial—it's imperative.

Vulnerabilities continue to surface in file transfer tools. In May, Australian cybersecurity firm Assetnote alerted Citrix to a critical vulnerability in the ShareFile storage zones controller, or SZC, in its cloud-based secure file-sharing and storage service known as Citrix Content Collaboration. Citrix patched the flaw on May 11, notified customers directly about the vulnerability and helped them lock it down. Citrix also blocked unpatched hosts from connecting to its cloud component, thus limiting any hacking impact to a customer's own environment. The U.S. Cybersecurity and Infrastructure Security Agency warned in August that the Citrix ShareFile vulnerability was being actively exploited by attackers. ... Security experts have warned users of secure file transfer software to safeguard themselves, given the risk of more such attacks perpetrated by Clop or copycats. One challenge with Clop's attacks is that the group has somehow continued to obtain access to zero-day vulnerabilities in the products, meaning even fully patched software could be - and was - exploited.

The analogy between fictitious responses produced by a machine and sensory phenomena in humans is clear: Both produce information that is not grounded in reality. Just as humans experiencing hallucinations may see vivid, realistic images or hear sounds reminiscent of real auditory phenomena, LLMs may produce information in their “minds” that appear real but is not. ... While the ultimate causes of AI hallucinations remain somewhat unclear, a number of potential explanations have emerged. These phenomena are often related to inadequate data provision during design and testing. If a limited amount of data is fed into the model at the outset, it will rely on that data to generate future output, even if the query is reliant on an understanding of a different type of data. This is known as overfitting, where the model is highly tuned to a certain type of data but incapable of adapting to new types of data. The generalizations learned by the model may be highly effective for the original data sets but not applicable to unrelated data sets.

This is likely your fault since you did not plan well and missed many things that became unexpected costs or delays. Also, there are known budget issues around migrating or developing new systems and how much they cost to operate after being deployed. We’re talking about both. Not everyone is an excellent planner, but there is a discipline to project management, including metrics and estimation approaches, that most IT projects choose to ignore. They provide a rough estimate of how long and how much money should be needed to do something meaningful in the cloud. Ignoring these guidelines is never good, so let’s learn from our mistakes and improve project planning. ... Engage in proactive communication with your cloud service providers to discuss your situation and explore any potential options for cost reduction. Yes, this means begging for a discount. Providers may offer flexible pricing plans, reserved instances, or cost optimization guides since it’s their system. Also, this may mean that you have to agree to future commitments for cloud service usage that may be out of this budget period. This could be an ethical or policy no-no at your company, so check with the CFO.

In addition to state-sponsored attacks by APTs, governments must deal with their fair share of criminal activity as well, particularly at lower levels of government where cybersecurity resources are especially scarce. This includes attacks against police departments, public schools, healthcare systems, and others. These attacks ramped up in 2023, a trend we expect to continue as cybercriminals look to easy targets from which to steal sensitive data like PII. Ransomware groups’ success is often less about technological sophistication and more about their ability to exploit the human element in cyber defenses. Unfortunately, this is exactly the area where we can expect AI to be of the greatest use to criminal gangs. Chatbots will continue to remove language barriers to crafting believable social engineering attacks, learn to communicate believably, and even lie to get what they want. As developers release ethically dubious and amoral large language models in the name of free speech and other justifications, these models will also be used to craft novel threats.