AM BEST: REGULATORY INITIATIVES, ECONOMIC EXPANSION DRIVE MALAYSIA’S NON-LIFE GROWTH

KUALA LUMPUR, Sept 10 (Bernama) -- AM Best has maintained a stable outlook on Malaysia’s non-life insurance segment, citing regulatory initiatives and economic expansion, which are driving robust premium growth and increasing insurance penetration.

The stable outlook on Malaysia’s non-life segment is also supported by de-tariffication of motor and fire insurance, as well as measures curbing medical inflation. However, it remains vulnerable to developments in the external environment.

The Best’s Market Segment Report, “Market Segment Outlook: Malaysia Non-Life Insurance”, stated that motor and fire insurance anchor the market, together accounting for more than 65 per cent of total non-life premiums.

In a statement, the global credit rating agency said pricing has progressively shifted towards a more risk-based approach since phased liberalisation of tariffs for these lines began in July 2016.

Over time, AM Best expects de-tariffication to drive product innovation, improve service quality, align pricing with underlying risks and enhance market efficiency, although it may pressure underwriting margins over the intermediate term.

“Malaysia’s non-life insurers continue to maintain healthy underwriting profits through disciplined underwriting and effective pricing strategies, supporting the industry’s long-term sustainability,” said AM Best senior financial analyst, Sin Yee Chuah, adding that the segment remains poised for continued growth.

The rating agency added that Malaysia’s non-life segment reported an improved underwriting profit in 2025, with a healthy combined ratio in the low-to-mid-90 per cent range, reflecting sustained underwriting discipline that supported profitability.

Malaysia’s economy continues to be supported by resilient domestic demand, particularly household consumption and investment, while strong demand for electrical and electronics exports and continued investment in data centres provide additional support.

The report also noted that a pilot phase of the recently introduced RESET Strategy, which introduced a standardised base medical and health insurance/takaful plan with a co-payment feature, is targeted for the second half of 2026, with full rollout expected by early 2027.

In addition, climate change is projected to increase the frequency and severity of extreme weather events, presenting floods as a persistent tail risk for insurers and exposing the segment’s profitability to greater volatility.

-- BERNAMA

GENERAL ROBOTICS ADVANCES ROBOT DEPLOYMENT WITH SELF-ENGINEERING GRID PLATFORM

 

GRID, the robot intelligence platform from General Robotics, auto‑engineers the full lifecycle of a complex pouring skill — from creation and testing to deployment and real‑world evaluation in a biolab.

KUALA LUMPUR, Sept 10 (Bernama) -- General Robotics, a company developing an intelligence layer for physical artificial intelligence (AI), has advanced its GRID robot intelligence platform to automate key stages of robot deployment, reducing the time and specialised expertise needed to bring robots into production.

The company in a statement said GRID can now use AI to automate processes ranging from robot onboarding and AI model integration to the creation and deployment of new robotic skills.

General Robotics founder and chief executive officer, Ashish Kapoor said the platform brings robotics knowledge spanning research, software, AI models, simulations, data and hardware into a single agent-first system, enabling intelligence to compound with each deployment.

The latest development has reduced robot onboarding time from one month to as little as two hours, while model ingestion has been cut from three days to as little as 20 minutes. Skill transfer across different robot form factors now takes as little as 1.5 hours, while new skill creation and deployment can be completed in as little as two days.

General Robotics said GRID uses knowledge graphs to convert each deployment, task, model and failure into structured, reusable intelligence while protecting customer data and intellectual property.

The platform supports robots from different manufacturers and form factors, allowing customers to select systems suited to specific tasks. Its growing ecosystem includes industrial robotics maker FANUC and bimanual mobile manipulation company Galaxea Dynamics.

The company said GRID customers include global companies in automotive manufacturing, port operations, energy generation and food and beverage production, as well as government agencies.

General Robotics said the robotics industry faces shortages of specialised talent and a fragmented development stack that requires bespoke integration of software, communications protocols and programming systems, limiting the transition from pilot projects to production-scale deployments.

GRID is designed to address these constraints through a modular library of robotic skills, foundation models and classical techniques that can be used across different robot types and applications.

The latest version of GRID can determine the skills required for a task, select and combine relevant models and approaches, create and run simulations, and deploy and monitor AI skills. The platform then uses real-world performance and failures to determine further refinements, creating a continuous development loop across connected robots and tasks.

-- BERNAMA

SHENZHEN FORUM ADVANCES ASIA-PACIFIC MEDIA COOPERATION

 

KUALA LUMPUR, Sept 7 (Bernama) -- The Asia-Pacific Media Forum in Shenzhen has brought together more than 400 representatives from over 220 media organisations, think tanks, government bodies, diplomatic missions, United Nations (UN) agencies and international organisations from 42 countries and regions.

Themed “Building a Path to Shared Prosperity for the Asia-Pacific Community: Media Consensus and Action”, the forum focused on strengthening regional media cooperation ahead of the 33rd APEC Economic Leaders’ Meeting, which China is scheduled to host later this year.

The event resulted in several new cooperation mechanisms, including the establishment of the Asia-Pacific Media Partnership Organizing Committee and the signing and exchange of media cooperation documents.

It also saw the launch of the “Gather in Shenzhen, Create a Better Asia-Pacific” Media Cooperation Initiative, which focuses on technology sharing, joint content production and professional training, as well as the release of the Shenzhen Consensus, according to a statement.

Under the technology pillar, Shenzhen plans to establish the Asia-Pacific Digital & Smart Media Intelligence Shenzhen Center to promote exchanges in digital media technologies, including artificial intelligence (AI) and the OpenHarmony ecosystem.

For content collaboration, the forum launched the “Witness Miracles, Seek Solutions” programme, which invites media organisations from across the region to jointly develop and broadcast reporting projects using Shenzhen as one area of focus.

The initiative also includes a Training Base for Asia-Pacific Youth Media Professionals, which is expected to host media professionals, particularly those under 45, for study visits, exchanges and training programmes.

The Shenzhen Consensus calls for greater regional cooperation in areas including digital transformation, AI, green development, media exchange and multilateral trade, while encouraging greater dialogue and mutual understanding across the Asia-Pacific.

-- BERNAMA

HAN’S LASER LEADS INTERNATIONAL SAFETY STANDARD FOR LASER PROCESSING EQUIPMENT

KUALA LUMPUR, Sept 3 (Bernama) -- Han’s Laser Smart Equipment Group Co Ltd announced that it led the development of ISO 11553-2:2026, the first international safety standard for complete laser processing equipment for hand-held or hand-operated machines developed under the ISO/IEC framework.

Han’s Laser in a statement said the international standard, ISO 11553-2:2026 Safety of machinery-Laser processing machines-Part 2: Safety requirements for hand-held or hand-operated laser processing machines, was officially released on Aug 28 after a seven-year development effort.

The release marks the first international safety standard for complete laser processing equipment under the ISO/IEC framework to be spearheaded by Han’s Laser, signifying a strategic transition for the company from a participant in standardisation to a standard-setter in the hand-held laser processing equipment sector.

The new standard advances the safety protection approach from “personnel management + end-user protection” to “inherently safe design measures embedded in the product itself”, encompassing risk assessment, safety functions, protective measures, control systems and information for use.

It specifies that primary responsibility for safety lies with the equipment’s inherent design rather than relying on operator training and personal protective equipment. The standard also provides a unified global safety benchmark for hand-held laser processing machines and serves as a technical foundation for market access.

With decades of technological expertise and practical innovation in intelligent laser equipment, Han’s Laser has accumulated extensive standardisation experience and brought together a broad network of experts.

During the development of ISO 11553-2:2026, an expert team from Han’s Laser convened more than 100 international meetings and successfully navigated multiple rigorous review procedures, including two IEC voting reviews, two ISO voting reviews, three CEN voting reviews and two EU Machinery Regulation compliance assessments.

Han’s Laser is a wholly-owned subsidiary of Han’s Laser Technology Industry Group Co Ltd. The company will continue to deepen its engagement in global laser standardisation efforts and contribute to the advancement of the international laser equipment industry in standardisation, safety and quality.

-- BERNAMA

Beyond Profit: IFTC Introduces Structured Approach to Measuring Trading Performance

 

^{IFTC officials and industry representatives gather on stage during the official inauguration, marking the introduction of a structured framework for transparent and risk-adjusted trading performance evaluation.}


KUALA LUMPUR, Aug 28 (Bernama) -- As financial trading becomes increasingly visible across digital platforms and online communities, the International Financial Trading Championship (IFTC) believes one fundamental question remains — how should trading performance be measured fairly?
 
A trader may show strong returns or a profitable account, but without understanding the level of risk undertaken, drawdown experienced and consistency of performance, headline returns alone may provide an incomplete picture of trading capability.
 
Building on its record recognition at the Malaysia, ASEAN and Asia levels, IFTC is seeking to address this challenge through a more structured approach to performance measurement.
 
Operating under the institutional governance of the International Financial Consultant Certified Institute (IFCCI), IFTC places measurable performance, risk discipline, transparency and consistent evaluation at the centre of its competition philosophy.
 
From Performance Claims to Measurable Evidence
 
The growth of online trading communities has made it easier than ever for traders to share results. However, performance is frequently communicated through screenshots, percentage returns or selected trading results, while the risks taken to achieve those results may receive less attention.
 
Two traders may generate similar returns while assuming significantly different levels of risk. Likewise, a high short-term return does not necessarily demonstrate whether a trader can maintain discipline and manage risk over a meaningful period.
 
IFCCI President Prof. Dato’ Dr. Kingston Chang said profitability remains important, but it should not be viewed in isolation.
 
“Profit is important, but profit is only one part of the story.
 
“If two traders generate the same return but one assumes significantly greater risk or experiences substantially higher drawdown, their performance should not necessarily be viewed in the same way.
 
“The industry needs a more balanced way of understanding trading capability — one that considers not only what was achieved, but also how it was achieved,” Chang said.
 
Introducing the Trader Performance Index
 
At the centre of IFTC’s approach is the Trader Performance Index (TPI), a performance measurement framework designed to evaluate trading results with greater consideration of risk.
 
Rather than ranking traders purely by absolute profit or percentage return, TPI considers the relationship between return and drawdown, providing a more balanced representation of trading performance.
 
The principle is straightforward: generating returns matters, but the amount of risk taken to generate those returns matters as well.
 
IFTC said no single measurement can fully define a trader’s ability. Instead, TPI provides a common performance reference that can be applied consistently within the IFTC framework, allowing eligible results to be compared using the same measurement principles.
 
Why a Minimum 30-Day Performance Period Matters
 
Time is another important element of the framework. Under IFTC’s latest structure, an eligible TPI performance record must cover a minimum period of 30 days before it can qualify for recognition within the wider IFTC ranking ecosystem.
 
The requirement is intended to reduce emphasis on isolated short-term results and provide a more meaningful period in which risk management, drawdown and trading behaviour can be observed.
 
IFTC Organising Chairman Kayden Chiew said the intention is to recognise demonstrated performance rather than momentary outcomes.
 
“A trader can have an exceptional day or week, but professional performance should be evaluated over a period that allows both return and risk behaviour to become more visible.
 
“The 30-day requirement provides a common minimum period and encourages participants to consider not only returns, but also how they manage performance throughout the competition,” Chiew said.
 
Same Measurement, Same Opportunity
 
IFTC believes a professional competition should create an environment where performance can speak for itself. Within such an environment, established traders and emerging participants should ultimately be evaluated against the same defined criteria.
 
Personal reputation, social media following or commercial influence should not determine performance ranking. Measurable results should. This philosophy forms part of IFTC’s broader effort to move trading competition from claims towards evidence, and from headline returns towards measurable performance.
 
Transparent rules and consistent competition requirements are intended to provide participants with greater clarity on how their performance is assessed. For IFTC, the purpose of a common performance framework extends beyond identifying winners. It also seeks to encourage a healthier understanding of trading performance, where risk management, discipline and consistency are recognised alongside profitability.
 
Chang said such standards are important to the longer-term professional development of the trading community.
 
“A credible trading ecosystem should not encourage the assumption that the highest return automatically represents the strongest trader.
 
“We want to encourage a culture where performance can be demonstrated, risk can be understood and results can be evaluated under common standards.
 
“From claims to evidence, and from returns to performance — that is the direction we believe professional trading competition should move towards,” he said.
 
Having established record recognition at the Malaysia, ASEAN and Asia levels, IFTC said the next stage of its development will apply these performance principles across a broader competition ecosystem.
 
Further details on how different competitions, industry partners and traders can participate within the IFTC framework will be introduced as part of its next phase.
 

SOURCE: International Financial Trading Championship (IFTC)

Sapient Intelligence Launches PRAXIST (Beta) to Accelerate a New Era of Autonomous AI-led Research and Development

• 49 gold-medal outcomes across 75 machine learning competitions from MLE-Bench achieving approximately US$3,000, versus 34 for Claude Code at US$38,000
• Rocket-landing accuracy reached 100% within 12 hours, demonstrating a proof-of-concept capability at Technology Readiness Level (TRL) 3

SINGAPORE, Aug 28 (Bernama-BUSINESS WIRE) -- Singapore-headquartered artificial general intelligence research company Sapient Intelligence today announced the launch of PRAXIST (Beta), an autonomous AI research and development (R&D) system that takes on complex technical problems and independently tests and validates potential solutions.

While AI has rapidly accelerated productivity in content generation and other well-defined workflows, R&D remains inherently difficult to automate and scale. Breakthroughs require testing multiple hypotheses, learning from successes and failures, and continuously determining which paths to pursue. Organizations must address this all while balancing specialist expertise, time, cost, and infrastructure.

A capacity multiplier

PRAXIST addresses this challenge as an R&D capacity multiplier. Rather than executing a predefined path, it independently explores which technical approach can best achieve a measurable objective. Users define the goal, parameters, and budget; PRAXIST then autonomously experiments and evaluates the strongest solution.

The approach has demonstrated promising results in controlled internal evaluations. On 75 challenging Kaggle competitions from MLE-Bench, a benchmark designed to test how well AI systems tackle complex, real-world machine learning problems across, PRAXIST achieved the highest-level result in 49 competitions at an approximate recorded model cost of US$3,000, compared with 34 highest-level results for Claude Code at approximately US$38,000 under the same evaluation conditions.

“Unlike general coding agents, which are built primarily to execute a specific task, PRAXIST is designed for long-horizon R&D, where the problem-solving approach itself may need to be discovered and adapted,” says William Chen, Co-Founder at Sapient Intelligence. “A conventional research team is ultimately constrained by the number of experiments its researchers can realistically run and evaluate. PRAXIST is designed to provide organizations with the ability to augment their existing teams with additional research capacity, enabling them to explore problems with a breadth and speed that would otherwise require significantly greater specialist resources.”

From autonomous research to cumulative scientific discovery

PRAXIST deploys multiple autonomous research peers to explore approaches in parallel, test hypotheses, investigate failures, and validate promising results. Unlike the tree-like search used by other similar systems, where each candidate inherits from one parent and weaker branches are pruned, PRAXIST uses a generation-layered research graph that preserves what every experiment teaches. This allows later generations to combine valuable mechanisms, evidence, and constraints across different lineages, including failed attempts. Rather than searching for the single best branch, PRAXIST constructs stronger solutions from discoveries made across branches.

Expansion and scale at speed

For organizations with limited AI or machine learning capabilities, PRAXIST can provide an AI research layer alongside existing domain expertise; for sophisticated teams, it can augment existing capabilities and increase the scale and breadth of R&D. A traditional robotics company, for example, can define problems, objectives, and constraints from an engineering perspective while PRAXIST conducts the AI research, experimentation, and validation needed to develop solutions, effectively serving as an in-house AI research team without requiring the company to build one.

Partner engineering results further illustrate this potential. In a partner-provided rocket-landing simulation, PRAXIST improved baseline to 100% within 12 hours, demonstrating a proof-of-concept capability at Technology Readiness Level (TRL) 3. In an industrial robotic SLAM problem, a partner team achieved 9.37cm of accumulated error after several months of development. PRAXIST reduced the error to 5.01cm within three days.

System application

PRAXIST can operate with proprietary data in private or customer-controlled environments, giving organizations greater control over their research infrastructure and sensitive intellectual property.

Initial applications are focused on sectors with intensive R&D requirements or highly measurable or simulatable optimization problems, including AI, engineering and robotics, manufacturing, finance, and health and drug discovery. Over time, Sapient Intelligence aims to extend PRAXIST beyond traditional R&D into broader business optimization, from inventory and sales to logistics and shipping.

“AI has mastered executing what we know. The next frontier is discovering what we don’t,” said Jin Li, Chief Scientist of PRAXIST. “PRAXIST is our first step toward making autonomous discovery a practical capability for organizations tackling complex problems. As we continue to develop the platform, we will look to expand beyond traditional R&D. Our ambition is to give organizations a fundamentally greater capacity to explore what is possible, enabling existing teams to pursue more experiments, approaches, and innovations while keeping human expertise and judgment at the center.”

About Sapient Intelligence

Founded and headquartered in Singapore, Sapient Intelligence is developing a new generation of AI systems that move beyond answering questions and executing predefined tasks to autonomously discover, reason, and optimize across complex problems without predefined solutions. By combining autonomous research systems with novel foundation-model architectures, Sapient Intelligence advances autonomous discovery through deeper reasoning, self-evolving capabilities, greater adaptability, and enhanced interpretability.

At the center of this vision is PRAXIST, the company’s flagship autoresearch system enabling breakthroughs in real-world R&D and business challenges. Sapient is also the creator of the Hierarchical Reasoning Model (HRM), a novel, brain-inspired foundation model architecture designed to enable deep reasoning with substantially greater training efficiency.

With a global team of more than 50 researchers and engineers operating across Singapore, Palo Alto, and Beijing, Sapient brings together experience from leading AI organizations and research institutions. The company connects foundational AI research with demanding applications in finance, healthcare, manufacturing, and logistics, developing original AI systems for enterprises, researchers, and institutions worldwide.

Learn more about PRAXIST here.

View source version on businesswire.com:
https://www.businesswire.com/news/home/20260827080331/en/

Contact

The Hoffman Agency on behalf of Sapient Intelligence
Sapient@hoffman.com

Source : Sapient Intelligence

--BERNAMA

LEARNING TREE INTRODUCES ANTHROPIC'S CLAUDE CERTIFICATION PREPARATION COURSES

KUALA LUMPUR, Aug 28 (Bernama) -- Learning Tree International has announced a new suite of live, instructor-led preparation courses for Anthropic’s Claude certifications, giving organisations a structured path to build role-based Claude expertise across business, developer, and architecture teams.

As organisations move from artificial intelligence (AI) experimentation to enterprise implementation, they need teams that can use AI tools responsibly, securely, and effectively in real business workflows.

In a statement, Learning Tree said Anthropic’s Claude has become a leading enterprise AI solution, recognised for advanced reasoning, large-context analysis, coding support and an emphasis on responsible AI.

Learning Tree’s Claude Certification Preparation courses help organisations develop validated, role-based expertise for business professionals, developers, and architects.

Aligned to Anthropic’s certification portfolio, the programmes combine exam preparation with practical instruction on designing, developing, governing, and operationalising Claude-powered solutions in enterprise environments.

The new courses support preparation for Anthropic’s four current role-based certifications, namely Claude Certified Associate – Foundations, Claude Certified Architect – Foundations, Claude Certified Architect – Professional and Claude Certified Developer – Foundations.

The courses include hands-on labs, enterprise-focused use cases, expert coaching and mock examinations, providing learners with practical experience applying Claude to business challenges while preparing for certification exams.

Learning Tree said Anthropic certification exams are currently available to eligible individuals employed by organisations in the Claude Partner Network, with final exam eligibility determined by Anthropic.

-- BERNAMA