World oil market looks well supplied in 2025, IEA says
In a monthly report, the IEA said world oil demand will rise by 720,000 barrels per day this year, down 20,000 bpd from last month's forecast, while supply will increase by 1.8 million bpd, up 200,000 bpd from last month's estimate.
"In the absence of a major disruption, oil markets in 2025 look well supplied," the agency said in a commentary on the market impact of Iran's conflict with Israel.
(Reporting by Alex Lawler and Enes Tunagur, Editing by Barbara Lewis)
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Khaleej Times
6 minutes ago
- Khaleej Times
As AI regulation tightens, legal foresight will define success
The clock is ticking for organisations that build, deploy, or rely on AI systems. With phase two of the European Union's AI Act fast approaching, the world's first binding AI regulation is ushering in a new era of enforceable AI governance. The act introduces strict requirements to ensure transparency, safety, and human oversight, grouping AI systems into four risk tiers that define how they are governed. Phase one, which came into force on February 2, 2025, banned AI systems deemed to pose unacceptable risk and required organisations operating in the EU to improve AI literacy among employees. The next phase places new obligations on general-purpose AI, including models such as ChatGPT, Claude, DALL·E, Midjourney, Anthropic, Meta, and Google BERT. These requirements mandate greater transparency, the creation of detailed technical documentation, and the disclosure of copyrighted materials used during training. European companies have already called for a pause on these measures, which the EU has rejected. From my perspective, this signals two realities. The first is that the era of enforceable AI regulation has arrived and will only continue to evolve. And the second, is that while responsible AI frameworks have been discussed for years, many businesses are still working out how to get prepared. Legal foresight as a strategic advantage Treating AI compliance as a reactive exercise has never been sustainable. But today, with the pace at which the regulatory landscape relating to AI and responsible and ethical AI, is evolving, legal foresight is more essential than ever. It must be embedded into how AI systems are built, deployed, and scaled from the outset. This is especially critical for general-purpose models, which are complex, fast-evolving, and capable of broad downstream impact. Integrating compliance into the development process, rather than retrofitting it later, helps reduce operational risk while strengthening trust with customers, partners, and regulators. This proactive approach is already being embraced in forward-looking jurisdictions such as the UAE, where regulation is positioned as a catalyst for innovation. In recent years, the UAE has formalised its AI governance through initiatives like the Artificial Intelligence and Advanced Technology Council (AIATC) and the UAE AI Charter, both of which emphasise ethics, transparency, and public interest. These efforts underscore the importance of co-design between government and industry to ensure that emerging technologies are not only cutting-edge, but also responsibly deployed. The co-regulation imperative No single organisation can navigate this evolving regulatory landscape alone. Engaging with industry peers, policymakers, and technology providers is essential. Co-regulation, where governments, private companies, and civil society collaborate to shape rules and guardrails which promote progress, is emerging as the most effective approach. The UAE's partnership with the World Economic Forum through the Global Regulatory Innovation Platform (GRIP) is a prime example. GRIP provides shared frameworks, benchmarking tools, and collaborative mechanisms that transform regulation from a constraint into an enabling model, aligning speed, inclusivity, and ethical standards. This collaborative mindset also shapes how Core42 designs its infrastructure. We embed regulatory and sovereignty requirements directly into our platforms to ensure compliance is not an afterthought. Our Sovereign Public Cloud, built on Microsoft Azure and supported by our sovereign controls platform Insight, addresses the complex challenges of data sovereignty and compliance for governments and regulated industries while still enabling public cloud innovation. With built-in sovereign control sets, comprehensive audit capabilities, and streamlined compliance management, organisations can manage sensitive data with confidence while benefiting from hyperscale agility and security. For customers requiring fully isolated environments for classified workloads, our Signature Private Cloud extends the same principles by combining the scalability of cloud with the control and assurance of on-premises systems. Preparing for the next frontier As governments worldwide move toward more comprehensive oversight through the regulation of AI, the organisations that thrive will be those that invest in proactive legal strategies today and seek to ensure their perspectives are tabled during the co regulation phase. In the Middle East, we are already seeing this next frontier take shape. The launch of the Responsible AI Foundation by G42 and Microsoft, in collaboration with MBZUAI, is establishing region-specific standards for fairness, transparency, and accountability. Additionally, G42's 2024 report on Sovereign AI Ecosystems highlights the growing importance of robust data governance frameworks as demand for data, storage, and computational power accelerates with AI advancements. As such, future regulations are likely to set clear parameters for data sovereignty, cross-border flows, and the accountability of AI systems operating on national infrastructure, addressing both technical integrity and the socio-cultural considerations unique to the region. In addition, continuous risk assessments, real-time auditing, and sector-specific governance will soon become standard practice. To prepare, legal teams must build cross-functional expertise, be creative and immersed in the dynamism of the environment in which the world is operating and work closely with engineers, product owners and data scientists to anticipate the regulatory landscape that will apply to what they are building rather than react to it. For companies developing or deploying AI, success will not hinge solely on technological breakthroughs but on the ability to navigate an increasingly complex legal environment. Those who approach this with foresight and integrity will be best positioned to earn public trust and capture long-term value.
Zawya
36 minutes ago
- Zawya
Lyft, China's Baidu to launch robotaxi service in Europe next year
U.S. ride-hailing company Lyft said on Monday it will partner with China's Baidu to deploy robotaxis across Europe starting next year, marking the Chinese search engine giant's first foray into the European self-driven taxi sector. The partnership will launch initially in Germany and the United Kingdom, with Baidu's electric RT6 robotaxis operating on Lyft's platform pending regulatory approval. The rollout aims to scale to thousands of vehicles across Europe in the coming years, leveraging Lyft's $200 million acquisition of European mobility app FreeNow, completed on Thursday. The deal gives Lyft access to operations in nine countries and over 180 cities, marking its first expansion outside North America. Under the partnership, Lyft will operate the platform, managing customer services and fleet logistics, while Baidu will supply autonomous vehicles and technical expertise. "What we're excited about with FreeNow is they have a deep, long-lasting relationship with regulators, and we want to go and have those conversations about how we do this," Lyft Executive Vice President of Driver Experience Jeremy Bird told Reuters. The UK government has fast-tracked its robotaxi timeline and is aiming for initial services with paying passengers by spring 2026. Companies such as Uber and Wayve are also preparing for deployment once regulatory clearance is granted. The deal comes as rival Uber has expanded its global robotaxi efforts, working with Waymo , WeRide, and Momenta, with a European launch also set for 2026. Uber's global robotaxi presence intensifies pressure on Lyft to expand beyond its North American home market and build its autonomous capabilities. Baidu's Apollo Go service currently operates over 1,000 vehicles across 15 cities globally and has completed more than 11 million rides, but its partnership with Lyft marks its first commercial move into Europe.
Tahawul Tech
an hour ago
- Tahawul Tech
Opinion: Veeam exec believes some organisations guilty of pulling the wool over their own eyes when it comes to data resilience
Tim Pfaelzer, Senior Vice President & General EMEA Manager at Veeam, has penned an op-ed, in which he states his belief that many organisations are guilting of fooling themselves when it comes to the strength of their own data resilience. For too long, business leaders have viewed their organization's data resilience from afar, relying on theoretical plans and a checklist mindset. This 2D perspective – where technical measures are simply ticked off a to-do list – fails to capture the full, real-world cross-organizational complexity of cyber threats. Ransomware, in particular, cannot be fully simulated on paper. This mentality has led to a dangerous false sense of security. Veeam research shows that more than 30% of organizations believe they are more resilient than they actually are. While they may have the right pieces in place, unless these elements work together in a rigorously tested, real-world incident response plan, they risk being exposed when a true crisis hits. With 69% of organizations having faced a ransomware threat in the past year, the time for blind confidence is over. Leaders must remove the wool from their eyes and take meaningful, proactive action. False Confidence, Real Consequences Data resilience can be deceptively complex, and gaps often remain hidden until it's too late. Many organizations fall into the trap of believing they are prepared, only to find out otherwise under attack. Of the organizations that fell victim to ransomware last year, 69% thought they were prepared beforehand. After experiencing an attack, confidence in their preparedness dropped by more than 20%. Although the majority of organizations had a ransomware playbook, less than half included essential technical components such as backup copies and containment or isolation plans. On the surface, everything may have appeared in order – but a closer inspection revealed significant vulnerabilities. The consequences of misplaced confidence are severe. Only 10.5% of organizations were able to successfully recover following a ransomware attack last year, leading to major business and operational impacts. The recent M&S ransomware incident is a high-profile example, causing not only service outages for customers but also an estimated £300 million hit to trading profits. The Evolving Threat Landscape Some organizations may have hoped that the disruption of major ransomware groups like BlackCat and LockBit by law enforcement would make the threat landscape easier to navigate. In reality, the threat has not diminished – it has evolved. Smaller groups and 'lone wolves' have quickly filled the gap, bringing new methods and tactics that further challenge organizational resilience. From 2D to 3D: The Path to True Resilience Regardless of how confident an organization may be in its data resilience, a deeper, more critical examination of its ransomware playbooks is essential. It is no longer safe to assume that what works on paper will hold up under real-life duress. Leaders must move from a flat, 2D perspective to a dynamic, 3D approach. Start with the big picture: Do you know what data you need to protect and where it resides? Are the key resilience measures, such as a predefined chain of command and regular backup verifications, in place? Drill down further: Are your security teams up to date on the latest attack trends? With 89% of organizations reporting their backup repositories targeted by threat actors, ensuring redundancy for your backups is now critical. Plugging the gaps is only the beginning. Organizations must stress-test their incident response plans with real-world simulations. It's not enough to rely on plan A – test plans B, C, D, and beyond, including scenarios where critical staff are unavailable or multiple crises occur simultaneously. This process often exposes blind spots that would go unnoticed in a theoretical plan. Turning Confidence Into Capability Leveraging frameworks like the Veeam Data Resilience Maturity Model (DRMM), developed in partnership with McKinsey, can help organizations move beyond blind confidence. Our findings show that organizations with a high degree of data maturity recover from ransomware incidents seven times faster than their less mature counterparts, and experience three times less downtime. By taking control of data resilience – grounded in rigorous testing, continuous improvement, and collective intelligence – organizations can replace blind confidence with real capability. In the current threat landscape, it's not a question of 'if' your organization will be attacked, but 'when'. The best time to prepare is now – because in data resilience, only true readiness will make the difference.
