Analysts believe that gold will not reach historical highs in the second half of 2026. This is because higher interest rates in the US, increased treasury bond yields, and a strengthening dollar are putting pressure on the precious metal, despite persistent geopolitical tensions that support demand for safe-haven assets.
Under current market conditions, gold is expected to trade within a limited price corridor until the end of the year. Any significant rise will depend on changes in monetary policy expectations or signs of economic slowdown, according to analysts.
Vijay Valecha, Director of Investments at Century Financial, noted that stronger-than-expected economic growth in the US and persistent inflation have led investors to anticipate a tighter Federal Reserve policy, which negatively affects the price of gold. He predicts that the precious metal will remain predominantly within current levels until the end of the year.
The CIO explained that if any major catalyst appears, such as a reversal in monetary policy expectations or an economic slowdown, gold could experience an upward trend reversal. Until then, fluctuation within the set range is expected.
Valecha also highlighted the inverse relationship between gold and the US dollar: a strengthening greenback usually lowers the price of bullion. Furthermore, higher treasury yields reduce the attractiveness of non-yielding assets like gold, increasing the opportunity cost of holding this metal.
Although geopolitical tensions continue to support demand for gold, some experts believe that the current conflict in the region has created a different dynamic compared to previous crises. Ole Hansen, Head of Commodity Strategy at Saxo Bank, stated that rising oil prices have increased inflation expectations, which has strengthened the US dollar and kept bond yields high, offsetting much of gold's traditional appeal as a safe haven.
According to Hansen, gold will trade in a wide range of $3900 to $4200 per ounce until inflation and interest rate issues become clearer. He added that attracting a new wave of investment demand will likely require a sustained breakthrough above the $4500 mark, as many investors are still awaiting greater clarity regarding the Middle East, inflation, and interest rate prospects.
Hansen also warned that until financing cost forecasts stabilize or decrease, non-yielding assets like gold are likely to face periodic headwinds.
Despite short-term difficulties, analysts remain optimistic about the long-term prospects for gold. Avad Issavi, Regional Press Secretary and Market Analyst at iFOREX, pointed out that continued central bank purchases, high government debt, and the need for portfolio diversification continue to support the precious metal.
However, he noted that achieving new record highs will likely require a combination of monetary policy easing, stable inflation, and persistent geopolitical uncertainty, rather than a single factor. Valecha also mentioned that in the UAE, higher prices are changing consumer behavior: consumers are increasingly opting for smaller weight jewelry or purchasing bullion and coins as a long-term investment, while investors continue to view gold as an important tool for portfolio diversification.
During the weekly steering committee meeting, where transformation results are reviewed, all indicators may appear positive: milestones have been met, training is complete, communications have taken place, the system is stable, and the dashboard is colored green. However, a critical question arises: have people actually started working differently?
Most CIOs are aware of this point. A program can confirm that technology has been implemented, employees have been trained, and users have logged into the system. But it cannot prove whether the organization has adopted the necessary behavioral models to realize the investments.
CIOs excel at determining the strategic significance of digital transformation: they secure funding, develop roadmaps, appoint strong program leaders, and implement governance. However, the strategy is passed on to operational managers, service executive directors, and line managers, who require fundamentally different skills for mobilization. This is where the subtle misalignment occurs.
While CIOs can track infrastructure performance in real time, monitor security threats, and study software usage in detail, human adoption remains difficult to observe.
Attendance at training, login data, and engagement metrics are often viewed as proof of implementation progress. These metrics are useful, but they speak more to familiarity with the changes than to whether the change itself has become ingrained. You can provide people with tools, training, and licenses, but you cannot explain what is preventing them from using the new system.
Something is hindering behavior, and without measuring this factor, one must guess. Thus, organizations fall into the trap of a false sense of confidence. Early adopters speak up actively, top performers compete, support the system, and report their readiness to management. Leaders striving to appear aligned demonstrate confidence, which is easily mistaken for organizational readiness, drowning out the voices of those who are silently stuck, confused, or uncertain.
The problem is not the inaccuracy of the metrics, but that they are being asked to prove what they were never designed to measure. The true cost of weak implementation often becomes apparent too late for cheap fixes. By the time performance data shows that expected benefits have not been realized, the implementation budget is spent, the delivery team has left, and workarounds have become part of the workflow. Only then does the organization begin to view implementation as a remediation problem, when it should have been managed as a live delivery risk.
A study cited by the Harvard Business Review showed that in 2022, the average employee faced 10 planned corporate changes, compared to two in 2016. Over the same period, employee readiness to support organizational change dropped from 74% to 43%.
This is often called resistance, but that explanation is too convenient. In many organizations, employees do not reject changes entirely; they simply do not have enough time and cognitive resources to learn, practice, and solidify one new way of working before the next initiative arrives. Colossal effort is expended convincing people of the vision they will gain. Far less effort is spent demonstrating how their daily work life will change.
Everyone is engaged when the transformation is conceptual and inspiring. But when reality sets in—what will change, what is expected of me, what am I losing—people feel overwhelmed, insecure, or even threatened. These feelings exist at a level of vulnerability that individuals cannot easily express because it might affect their perceived performance.
The result is participation without adoption. People attend sessions, complete required tasks, and log in when prompted. Then, when operational pressure returns, they revert to the process they already trust. The program registers completion, but the business continues to function around the new technology, not through it.
Weak implementation rarely stops at one program. Workarounds and tricks developed by people become part of business operations. The next initiative is forced to compete with these habits, as well as any capability gaps left by the previous program.
Before moving to enterprise software, I worked for many years in industrial automation. We could precisely measure machine performance. We knew when equipment was operating within tolerance, when quality and performance were deteriorating, and when intervention was required. But we could not measure as well whether people had changed their way of working around these machines.
Enterprise technologies suffer from the same blind spot. Ask a CIO if their business is technically ready for launch, and they will answer accurately. Ask if their people are ready, and the answer will be subjective—each manager looks through the lens of their own activities and says what they believe to be true, not what they know to be true.
In practice, closing this gap means breaking down changes into smaller, role-based actions and monitoring the development of these behavioral models over time. At Change Logic, we analyze this across three dimensions: participation, ownership, and confidence.
Participation shows whether a person is engaging in the changes. Ownership shows whether the new behavior is absorbed and performed without constant external intervention. Confidence indicates whether a person believes in their ability to act independently and whether existing evidence supports this belief. This process allows leaders to identify different types of risks: one employee may not participate at all; another may participate but fail to take ownership; a third may report high confidence while continuing to make mistakes. A single completion score hides these differences. Behavioral intelligence makes them actionable.
For CIOs, this means the ability to see where adoption is forming, where it remains fragile, and where it has stalled while the program is still active, allowing for cost-effective course correction. Organizations are already successful at making technology observable. The next advantage will come from applying the same discipline to behavioral change. Until leaders can see if people are working differently, the green transformation dashboard will remain an assumption, not a fact.
AutoFlight is accelerating the commercialization of low-altitude logistics by collaborating with SF Express in marine wind energy, obtaining certification in Indonesia, and utilizing a priority cargo transport strategy. The goal is to commence real operational routes by 2027.
AutoFlight's technologies are transforming from a stage of conceptual competition to practical operation. The partnership with SF Express for wind farm logistics at sea represents a tangible step toward commercial low-altitude operations. The company has focused on maritime scenarios because they solve real problems: traditional ship transport takes 3 to 5 hours one way, whereas eVTOL reduces travel time to 15–20 minutes.
This choice aligns with the company's strategy, which has always prioritized cargo over passengers, as well as the national development plan that prioritizes freight transport first, followed by isolated areas, and then urban centers.
The selection of the marine environment was deliberate, as conditions with a constant Force 7 wind during takeoff and landing allow for the stress testing necessary to obtain certification. The zero-emission SUNSHIP water port integrates a takeoff platform, intelligent dispatching, and green energy equipment that can be deployed within hours on rivers, lakes, and seas.
According to AutoFlight, the marine energy industry will require over 100 SUNSHIP units and more than 300 eVTOL aircraft. The partnership with SF Express will ensure the supply of a complete set of equipment for marine energy, island logistics, and maritime applications. Planned tests and inspections have begun immediately, with the first delivery expected within a year.
The V2000CG eVTOL model is unique in its tonnage class and possesses all three types of Chinese certification. It has received the world's first overseas eVTOL certificate from Indonesia. The Shengshilong passenger aircraft, with a range of 200 km, aims to obtain certification by 2027.
International expansion is gaining momentum. In addition to Indonesia, AutoFlight is pursuing parallel certification in Japan, South Korea, and the Middle East, accumulating experience working within various regulatory systems. Indonesia's geography, consisting of islands, creates a natural demand for inter-island transport of fresh produce, medicine, and emergency supplies, positioning the V2000CG as a revolutionary logistics solution for Southeast Asian archipelagos. The company plans to establish several standardized operational routes by the end of 2026 or early 2027.
The sector has progressed from being a new growth engine in 2024 to an emerging industry in 2025, and by 2026, it has become a key national industry alongside semiconductors, aerospace, and biomedicine. AutoFlight Senior Vice President Se Jia emphasized that the next five years will be characterized by the prosperity of application, rather than mass universal operations, with standardization of application in landmark scenarios becoming the norm. The industry problem has shifted from whether the technology can fly to whether operations can become normalized and profitable.
Safety remains the main limiting factor, requiring the accumulation of flight hours and safety records through meticulous management of airworthiness, air traffic control, and operator management. Once public trust in low-altitude transport is established, the technological maturity of the infrastructure is already sufficient for rapid large-scale deployment, ensuring the global competitiveness of China's low-level economy due to its complete new energy industry chain, mature aviation manufacturing base, and the richest application scenarios in the world.
The embodied intelligence robotics industry in China has transitioned from the pilot testing stage to a critical period of scalable commercialization. This is evidenced by the first PwC China report on intelligent robotics development, which was presented on July 22 at the WAIC 2026 event, coinciding with the official establishment of the PwC China Artificial Intelligence Research Institute.
Morgan Stanley separately forecasts that the humanoid robot market in China will reach $2 billion in 2026 and grow to $15 billion by 2030. Furthermore, they significantly increased the forecast for unit shipments in 2026 from 28,000 to 50,000 units, reflecting accelerated commercial adoption.
The white paper identifies three distinct product forms by 2030. Industrial robotic manipulators will dominate standardized operations in fixed locations within the semiconductor, automotive, and high-tech industries. Wheeled and quadruped robots, utilizing efficient mobility and terrain adaptability, will expand their applications in warehouse inspection, restaurant delivery, and mobile scenarios, becoming the primary means of near-term commercial deployment.
Humanoid robots, adapted to human infrastructure and tools, will unlock unique value in elderly care, flexible assembly, and specialized environments. However, the report cautions that no single form of humanoid robot will replace all scenarios within 3–4 years, and over-concentration on them could lead to inefficient resource allocation.
On the supply side, players from different categories are converging. Robotics startups are pioneering the mass production of humanoid robots. Established robot manufacturers are leveraging accumulated technical expertise as industry benchmarks. Giants of traditional industrial automation are entering the field due to advantages in core components and manufacturing. Automotive and internet companies are participating, relying on the scale of industrialization and AI algorithm prospects, respectively.
Three strategic assets of China provide the most comprehensive industrial support system globally, facilitating a rapid iterative component cycle, dual reinforcement of policy and capital supporting scaled investment, and massive production scenario data that creates an ideal commercial environment for accelerated industry development.
Key challenges remain in the form of dual constraints of model data and supply-demand. The demand for computing power is growing exponentially, and linear capacity expansion cannot resolve the widening gap between supply and demand. Within the four scaling laws—covering pre-training, post-training, testing, and agent scaling—the supermarket of training and inference is developing interdependently. Reinforcement learning across clusters has become a new pillar for intelligence scaling, requiring heterogeneous computational support at an unprecedented scale.
The model-level strategy involves using vertical production processes in the short term to create scenario-specific competitive advantages while gradually building data collection systems and iterating to close the gap with global leaders. Prospects are supported by hybrid quadruped robots from Kepler Robotics and humanoid products demonstrated at WAIC, as well as updates to Kepler K3 and intensive iteration visible among major players. As PwC Strategy Partner Lin Jiund said, the industry has completed technology and scenario validation and entered the phase of scalable commercialization.
