Daily Tech Digest - January 31, 2026


Quote for the day:

"The most difficult thing is the decision to act, the rest is merely tenacity." -- Amelia Earhart



Security work keeps expanding, even with AI in the mix

Teams with established policies report greater confidence that AI outputs pass through review steps or guardrails before influencing decisions. Governance work spans data handling, access management, auditability, and lifecycle oversight for AI models and integrations. Security and compliance considerations also affect how quickly teams operationalize automation. Concerns around data protection, regulatory obligations, tool integration, and staff readiness continue to influence adoption patterns. Budget limits and legacy systems remain common constraints, reinforcing the need for governance structures that support day-to-day execution. ... Teams managing large tool inventories report higher strain, particularly when workflows require frequent context switching. Leaders increasingly view automation and tooling improvements as key levers for retaining staff. Practitioners consistently place work-life balance and meaningful impact at the center of retention decisions. ... Many teams express interest in workflow platforms that connect automation, AI, and human review within a single operational layer. These approaches focus on moving work across systems without constant manual handoffs. Respondents associate connected workflows with higher productivity, faster response times, improved data accuracy, and stronger compliance tracking. Interoperability also plays a growing role. Security teams increasingly consider standardized frameworks and APIs that allow AI systems to interact with tools under controlled conditions. 


Human risk management: CISOs’ solution to the security awareness training paradox

Despite regulatory compliance requirements and significant investment, SAT seems to deliver marginal benefits. Clearly, SAT is broken — even with peripheral improvements like synthetic phishing tools. So, what’s needed? Over the next few years, organizations should shift from static/sporadic security training to an emerging discipline called human risk management (HRM). ... HRM is defined as a cybersecurity strategy that identifies, measures, and reduces the risks caused by human behavior. Simply stated, security awareness training is about what employees know; HRM is about what they do. To be more specific, HRM integrates into email security tools, web gateways, and identity and access management (IAM) systems to identify human vulnerabilities. Furthermore, it measures risk using behavioral data and pinpoints an organization’s riskiest users. HRM then seeks to mitigate these risks by applying targeted interventions such as micro-learning, simulations, or automated security controls. Finally, HRM monitors behavioral changes so organizations can track progress. ... From an ROI perspective, HRM offers a much more granular approach to cyber-risk mitigation than standard SAT. CISOs and HR managers can report on improved cyber hygiene and behavior, rather than how many employees have been trained and past generic tests. Repeat offenders are not only identified but also provided with personalized training tools and attention. Ultimately, HRM makes it possible to show a direct correlation between training and a reduction in actual security incidents. ...


The Human Exploit: Why Wizer Is the Secret Weapon in the War for Your Digital Soul

We are currently witnessing a systemic failure in how we prepare people for a digital world. From the moment a child gets their first school-issued tablet to the day a retiree checks their pension balance, every individual is a target. This isn’t just a corporate problem; it’s a societal one. That is why I’ve been following the rise of Wizer, a firm that has cracked the code on making security training not just tolerable, but actually effective. ... It is no coincidence that the financial industry has become Wizer’s most aggressive adopter. In banking, trust is the only product you’re actually selling. If a customer’s account is drained because an employee fell for a “vishing” attack—where a hacker samples an IT person’s voice from a voicemail to impersonate them—the damage to the brand is catastrophic. Financial institutions are currently the biggest fans of the platform because they operate under a microscope of regulation and extreme risk. They realized early on that a 45-minute annual compliance video is a waste of time. Wizer’s approach is different; it feels more like an app—specifically Duolingo—than a corporate lecture. ... One of the most profound insights Gabriel Friedlander brings to the table is the necessity of the “Security Awareness Manager” (SAM). Historically, security training was a secondary task for a stressed-out IT admin who would rather be configuring a server. That is a recipe for failure. To build a true culture of security, you need a dedicated facilitator.


Chinese APTs Hacking Asian Orgs With High-End Malware

A pile of evidence suggests that this campaign was carried out by a Chinese APT, but exactly which is unclear. Chinese threat actors are notorious for sharing tools, techniques, and infrastructure. Trend Micro found that this one — which it currently tracks as Shadow-Void-044 — used a C2 domain previously used by UNC3569. A Cobalt Strike sample on one of its servers was signed with a stolen certificate also spotted in a Bronze University campaign. And they linked one of its backdoors to a backdoor developed by a group called "TheWizards," not to be confused with the equally maligned basketball team. A second, separate threat actor has also been using PeckBirdy since at least July 2024. With low confidence, Trend Micro's report linked the group it labeled Shadow-Earth-045 to the one it tracks as Earth Baxia. This campaign was more diverse in its methods, and its targeting, involving both Asian private organizations and government entities. Chinese APTs habitually perform cyberespionage against government agencies in the APAC region and beyond. Trend Micro tells Dark Reading, "These two campaigns remind us that the boundary between cybercrime and cyberespionage is increasingly blurred. One tool used in different [kinds of] attacks is [becoming] more and more popular."



AI agent evaluations: The hidden cost of deployment

Agent evals can be complicated because they test for several possible metrics, including agent reasoning, execution, data leakage, response tone, privacy, and even moral alignment, according to AI experts. ... Most IT leaders budget for obvious costs — including compute time, API calls, and engineering hours — but miss the cost of human judgment in defining what Ferguson calls the “ground truth.” “When evaluating whether an agent properly handled a customer query or drafted an appropriate response, you need domain experts to manually grade outputs and achieve consensus on what ‘correct’ looks like,” he adds. “This human calibration layer is expensive and often overlooked.” ... The sticker shock of agent evals rarely comes from the compute costs of the agent itself, but from the “non-deterministic multiplier” of testing, adds Chengyu “Cay” Zhang, founding software engineer at voice AI vendor Redcar.ai. He compares training agents to training new employees, with both having moods. “You can’t just test a prompt once; you have to test it 50 times across different scenarios to see if the agent holds up or if it hallucinates,” he says. “Every time you tweak a prompt or swap a model, you aren’t just running one test; you’re rerunning thousands of simulations.” ... If an organization wants to save money, the better alternative is to narrow the agent’s scope, instead of cutting back on testing, Zhang adds. “If you skip the expensive steps — like human review or red-teaming — you’re relying entirely on probability,” he says.


Social Engineering Hackers Target Okta Single Sign On

What makes these attacks unusual is how criminals engage in real-time conversations as part of their trickery, using the latest generation of highly automated phishing toolkits, which enable them to redirect users to real-looking log-in screens as part of a highly orchestrated attack. "This isn't a standard automated spray-and-pray attack; it is a human-led, high-interaction voice phishing - 'vishing' - operation designed to bypass even hardened multifactor authentication setups," said threat intelligence firm Silent Push. The "live phishing panel" tools being used enable "a human attacker to sit in the middle of a login session, intercepting credentials and MFA tokens in real time to gain immediate, persistent access to corporate dashboards," it said. Callers appear to be using scripts designed to walk victims through an attacker-designated list of desired actions. ... At least so far, the campaign appears to center only on Okta-using organizations. ShinyHunters and similar groups have previously targeted a variety of SSO providers, meaning hackers' focus may well expand, Pilling said. The single best defense against live phishing attacks that don't exploit any flaws or vulnerabilities in vendors' software, is strong MFA. "We strongly recommend moving toward phishing-resistant MFA, such as FIDO2 security keys or passkeys where possible, as these protections are resistant to social engineering attacks in ways that push-based or SMS authentication are not," Mandiant's Carmakal said.


AI agents can talk to each other — they just can't think together yet

Current protocols handle the mechanics of agent communication — MCP, A2A, and Outshift's AGNTCY, which it donated to the Linux Foundation, let agents discover tools and exchange messages. But these operate at what Pandey calls the "connectivity and identification layer." They handle syntax, not semantics. The missing piece is shared context and intent. An agent completing a task knows what it's doing and why, but that reasoning isn't transmitted when it hands off to another agent. Each agent interprets goals independently, which means coordination requires constant clarification and learned insights stay siloed. For agents to move from communication to collaboration, they need to share three things, according to Outshift: pattern recognition across datasets, causal relationships between actions, and explicit goal states. "Without shared intent and shared context, AI agents remain semantically isolated. They are capable individually, but goals get interpreted differently; coordination burns cycles, and nothing compounds. One agent learns something valuable, but the rest of the multi-agent-human organization still starts from scratch," Outshift said in a paper. Outshift said the industry needs "open, interoperable, enterprise-grade agentic systems that semantically collaborate" and proposes a new architecture it calls the "Internet of Cognition," where multi-agent environments work within a shared system.


Building Software Organisations Where People Can Thrive

Trust builds over time through small interactions. When people know what to expect and how to interact with each other in tough moments, trust is formed, Card argued. Once trust is embedded, teams are more likely to take risks by putting themselves out there to be wrong and fail fast, and that is where the magic happens. You need to actively address bias and microaggressions. If left unchallenged, they quietly erode trust and belonging. Being proactive, fair, and consistent in addressing these behaviours signals your values clearly to the wider organisation, Card said. At the heart of it all is the belief that people-first leadership is performance leadership, Card said. When we take the time to build inclusive, resilient cultures, success follows, not just for the business, but for everyone within it, he concluded. ... psychological safety is the next level up from a trusting environment. Both are the foundations of any healthy, high-performing culture. Without them, people hold back; they’re less likely to share ideas, admit mistakes, or challenge the status quo. And that means your team won’t grow, innovate, or build strong relationships. If you want to build a culture that lasts, where people thrive, not just survive, then building trust and safety isn’t optional. It has to be intentional. And once it’s in place, it unlocks everything else: collaboration, resilience, accountability, and growth.


Stop Delivering Change, Start Designing a Business That Can Actually Grow

Legacy and emerging technologies sit side by side, often competing for attention and investment. Manual and systemised processes overlap in ways that only make sense to the people living inside them. Long-standing roles carry deep, tacit knowledge, while new-in-career roles arrive with different expectations, skills, and assumptions about how work should flow. Each layer is changing, but rarely in a deliberate, joined-up way. When leaders do not have a shared, design-level understanding of how these layers interact, decisions are made in isolation. ... Programme milestones become a proxy for progress. Technology capability becomes a proxy for readiness. Productivity targets replace understanding. Designing the next-generation business model requires a different kind of insight—one that shows how people, process, data, and technology interact end to end. One that makes visible where human judgement still matters, where automation genuinely adds value, and where the handoffs between the two are quietly breaking down. ... Growth and productivity are not things you add through execution. They are the result of deliberate design choices. A business model fit for today makes explicit decisions about what is standardised and what is differentiated, what is automated and what is augmented, what relies on experience and what demands new capability. Those decisions cannot be delegated to programmes alone. They sit squarely with leadership.


Beyond Human-in-the-Loop: Why Data Governance Must Be a System’s Property

The reliance on human action creates a false sense of control; although governance artifacts do exist, responsibility for accountability exists outside the formal governance system and is therefore difficult to enforce. ... The dominant structures of governance today represent a human-in-the-loop system model. In a human-in-the-loop model, technology is used primarily to automate the completion or execution of specific tasks, such as the movement of data between systems, checking the validity of data between systems, and enhancing data that has been provided or created by other systems. The responsibility for the outcome of an automated governance system is not part of the automated system itself. Therefore, humans have the ability to resolve disputes between systems, approve any exceptions made by systems, and determine what is true when using different systems produces different conclusions. ... As data ecosystems continue to expand, we see recurring patterns of failure emerge. As a result, stewardship teams tend to create bottlenecks in their processes, as the volumes of existing exceptions continue to grow much faster than the capability to resolve those exceptions. The presence of escalation paths creates delays in decision-making processes, leading to inconsistencies in the products or services being delivered. Over time, informal methods of addressing issues become accepted as standard operating procedures. Controls within the organization are bypassed in an effort to deliver projects on time.

No comments:

Post a Comment