Breaking AI & Tech News Analyzed

The Leadership Shift at Anthropic Andrej Karpathy, the AI researcher who co-founded and formerly worked at OpenAI and previously led AI at Tesla, has joined Anthropic. Karpathy announced his move on X, stating that he is excited to join the team and get back to R&D.; Karpathy's Role in Pre-Training Karpathy started this week at Anthropic, where he is working on pre-training under team lead Nick Joseph. Pre-training is responsible for the large-scale training runs that give Claude its core knowledge and capabilities. Karpathy will start a team focused on using Claude to accelerate pre-training research. The Significance of Karpathy's Move Karpathy is one of the few researchers who can bridge the gap between LLM theory and large-scale training practice. Tapping him to build such a team is a clear sign from Anthropic that it believes AI-assisted research, rather than pure compute, is how it stays competitive with OpenAI and Google. Karpathy's Background Co-founded OpenAI and worked on deep learning and computer vision until 2017 Led Tesla's Full Self-Driving (FSD) and Autopilot programs from 2017 to 2022 Returned to OpenAI for one year before leaving in 2024 to start Eureka Labs, a startup dedicated to applying AI assistants to education Anthropic's Recent Hires Anthropic has also brought on Chris Rohlf to its frontier red team, which stress-tests advanced AI models against severe threats. Rohlf is a veteran of the cybersecurity industry with more than 20 years of experience. The Future of AI Research Karpathy's move to Anthropic and the company's focus on AI-assisted research signal a new direction in the AI landscape. As Karpathy stated, "I think the next few years at the frontier of LLMs will be especially formative."

Israel and Argentina launched a direct, twice‑weekly El Al flight in November, positioning the route as a political bridge rather than a profit‑center. The service, spanning 12,000 km and lasting about 16.5 hours, is intended to cement Israel’s foothold in Latin America while offering a logistical bypass for officials facing European legal scrutiny.The Launch of the Tel Aviv‑Buenos Aires Direct FlightThe inaugural flight was announced by Prime Minister Benjamin Netanyahu and Argentine President Javier Milei during a ceremony in East Jerusalem, underscoring a shared ideological alignment. The route is promoted as the first tangible step of the Isaac Accords, a Latin‑American framework modelled on the Abraham Accords.Frequency: twice a week (Tuesdays and Saturdays)Distance: 12,000 km (7,460 mi)Duration: 16.5 hours, the longest El Al route to dateSubsidy: 20 million shekels (≈$5.4 m) over three yearsFinancial and Operational Numbers Behind the RouteEl Al’s booking launch on May 7, 2026 revealed modest commercial interest. In 2025, passenger traffic between the two nations reached roughly 55,300 travelers—a 37 % rise from 2024 but still below the pre‑pandemic peak of 71,200 in 2019.The long‑haul flight incurs higher operating costs because Israeli aircraft are barred from several African airspaces, forcing a costly detour over the Mediterranean and Atlantic.Geopolitical Implications for Israel’s Latin American StrategyAnalysts such as Ihab Jabarin argue the flight is a “permanent corridor for security and tech businessmen,” allowing Israel to export cybersecurity, AI, and surveillance expertise to Latin American partners. The route also provides a safe travel channel that sidesteps European arrest warrants for Israeli officials implicated in the Gaza conflict.By aligning with right‑wing leaders like Milei, Israel seeks to showcase ideological allies in a region where left‑leaning governments (e.g., Brazil’s Lula) have condemned its actions.What the Flight Signals for Future Israel‑Latin America RelationsIf passenger demand stabilises, the Tel Aviv‑Buenos Aires link could become a template for similar routes to other Latin American capitals, deepening cooperation in security, counter‑terrorism, and artificial intelligence.However, domestic pushback in both countries—ranging from economic concerns in Israel to accusations of “imperialist war” in Argentina—could limit expansion. The success of the service will hinge on the Argentine Jewish community’s support and the ability to mitigate logistical costs.

Standard Chartered announced a plan to cut more than 7,000 jobs over the next four years, driven by the bank’s expanding use of artificial intelligence. Chief executive Bill Winters framed the reduction as a shift from lower‑value human capital to financial and investment capital.AI‑Driven Workforce Reduction Plan UnveiledThe London‑headquartered lender said it will remove roughly 15% of its back‑office roles by 2030, targeting about 7,800 redundancies out of a back‑office headcount of more than 52,000. The cuts are positioned alongside higher shareholder‑return targets in a strategy update aimed at cementing profitability.Back‑Office Redundancies Targeted Across Global HubsThe most affected centres are located in Chennai, Bengaluru, Kuala Lumpur and Warsaw, where routine processing functions are slated for automation and AI‑enabled redesign.Numbers Behind the Cuts: 7,800 Redundancies and $190 million Provision7,800 back‑office jobs to be cut (≈15% of that segment).Back‑office workforce: > 52,000 employees.Total global staff: nearly 82,000.Precautionary provision for Middle East conflict: $190 million (£142 million) in the first quarter.Strategic Implications for StanChart and the Banking SectorThe restructuring underscores a broader industry trend where major banks leverage AI to streamline operations, curb costs, and counter rising cyber‑threats. By positioning AI as a “huge facilitator and enabler,” StanChart aims to transition from a potential takeover target to a sustainably profitable lender, while also addressing succession‑planning concerns surrounding Bill Winters’s long tenure.Future Outlook: AI Integration and Market ResilienceAnalysts expect continued AI deployment to shape staffing models across global banks, potentially prompting further efficiency‑driven reductions. Despite geopolitical headwinds—such as the ongoing Iran conflict that could force Asia‑Pacific banks to raise loan‑loss provisions—StanChart’s leadership asserts the institution remains “extremely resilient” and poised to meet its growth targets.

Executive Summary: Iran's Hormuz Cable Fee ProposalIran has floated a plan to charge US tech companies for using the undersea internet cables that pass through the Strait of Hormuz. The proposal, aired by state‑linked outlets Tasnim and Fars, claims the scheme could generate hundreds of millions of dollars each year, but experts question its legality and technical feasibility.Details of the Proposed Licence RegimeThe media brief outlines three core elements:Impose licence fees on foreign firms that transmit data over the subsea cables.Require the so‑called “technology giants” – specifically Meta, Google, Amazon and Microsoft – to operate under Iranian law, effectively forcing joint‑venture arrangements.Monopolise repair and maintenance services for the cables, charging the world for any restoration work.Iran justifies the move by citing article 34 of the 1982 UN Convention on the Law of the Sea, which it interprets as granting rights over the seabed of the strait.Financial Estimates and Comparative BenchmarksWhile the exact figure is vague, Tasnim suggests the scheme could bring in hundreds of millions annually. For context, the proposal references Egypt’s model, where fees on cables crossing Egyptian territory are estimated to generate between $250 million and $400 million per year, though precise revenues are not publicly disclosed.Strategic and Operational Implications for the Gulf RegionSeven major cables run beneath the Hormuz strait, many supporting the rapid AI and cloud expansion in Gulf states. Potential consequences include:Disruption of regional internet traffic if fees are enforced or if repair ships are deterred.Limited global impact, as most traffic on these cables serves Gulf countries rather than trans‑Eurasian routes.Increased geopolitical tension, especially given US naval patrols and the strategic importance of the waterway.Experts note that most cables do not terminate in Iran, making fee collection technically challenging. Additionally, imposing tolls would likely require threats or physical interference, a step not previously observed.Outlook: Feasibility, Enforcement, and Regional TensionLegal analysts highlight sanctions and international law as major obstacles. Technically, separating traffic by company is infeasible, and cutting or seizing cables would demand capabilities Iran does not demonstrably possess. Even if Iran attempted to threaten repair vessels, such ships typically avoid operating under fire, potentially prolonging any disruption.In the near term, the proposal appears more rhetorical than actionable, serving as a bargaining chip in the broader US‑Iran confrontation. Unless Iran can develop the requisite maritime and cyber‑monitoring infrastructure, the likelihood of a sustained, enforceable fee regime remains low.

Starmer’s Call for a Closer EU PartnershipKeir Starmer announced that the United Kingdom should pursue a tighter relationship with the European Union, emphasizing shared interests in trade, security and climate policy.Details of the Remarks and Their Immediate ContextDate of statement: 18 May 2026Venue: televised interview with the GuardianKey quote: “We want a partnership that works for both sides, not a debate about re‑joining.”Background: Labour’s election manifesto calls for “closer ties” but stops short of a full EU membership pledge.Financial Context Lacks Concrete NumbersThe speech did not include specific fiscal projections, leaving the economic impact of deeper cooperation open to interpretation. Analysts note that without quantified trade gains or cost estimates, the policy’s budgetary implications remain speculative.Political and Trade Ramifications for BritainPotential easing of customs frictions with the EU.Strengthening of security collaboration on counter‑terrorism and cyber‑defence.Possible friction within the Conservative opposition, which may portray the stance as a soft‑Brexit.Domestic debate over sovereignty versus economic pragmatism.Outlook for UK‑EU Relations Under a Labour GovernmentIf Labour wins the next general election, the expectation is a gradual alignment with EU standards in areas such as climate regulation and data protection, while maintaining the UK’s sovereign status. The next 12‑month horizon will likely see formal negotiations on sector‑specific agreements rather than a full membership discussion.

Anthropic’s Claude Mythos to be Presented to the Financial Stability BoardAnthropic will brief the Financial Stability Board (FSB), chaired by Bank of England governor Andrew Bailey, on the cyber‑defence implications of its Claude Mythos model, which has raised alarm among security experts.Mythos is not being released publicly; access is limited to select tech firms and banks such as Apple and JP Morgan.The briefing follows a report by the Financial Times and confirmation from a source familiar with the discussions.The FSB’s membership includes senior officials from the US, UK, Australia and China.Quantifying Mythos’ New Cyber‑Testing PerformanceThe UK’s AI Security Institute (AISI) noted a “notable capability jump” in the version shown to banks. In the “cooling tower” test, Mythos succeeded in 3 out of 10 attempts – a first for any model evaluated by AISI.Previous iterations had not completed the test.AISI reports that the length of autonomous cyber tasks has doubled within months.Implications for Global Financial CybersecurityThe briefing comes as the International Monetary Fund (IMF) warned that AI‑driven cyber risks are rising for financial stability. Central bank leaders, including Goldman Sachs CEO David Solomon and JP Morgan CEO Jamie Dimon, have already expressed heightened awareness of Mythos’ capabilities.Cyber risk does not respect borders; inconsistent oversight could weaken the interconnected financial system.Experts caution that most breaches still stem from traditional weaknesses such as weak authentication.What the Next Phase of AI‑Driven Cyber Risk May Look LikeAISI is developing tougher hacking tests to track AI progress, while the FSB is expected to issue recommendations for coordinated oversight among regulators. If the trend of rapid capability gains continues, financial institutions may need to embed AI‑specific cyber‑defence measures into their risk frameworks.Potential for tighter collaboration between AI developers and regulators.Increased scrutiny of AI models before deployment in critical infrastructure.

The Growing Threat of Business Crime in the UKUK business leaders are issuing a stark warning that crime has become an increasingly "serious barrier" to growing Britain's economy, with two-fifths of companies experiencing some form of criminal activity in the past year. The British Chambers of Commerce (BCC) is calling on the government to provide "a step change in the support businesses can count on" as businesses face rising levels of theft, fraud, and cyber-attacks.Rising Crime Statistics Across Business SectorsThe BCC's research, based on a survey of 1,411 firms, reveals that crime against businesses is widespread and growing. Key findings include:Two-fifths of companies experienced some form of crime in the past yearOne-fifth of companies faced fraud or scams21% experienced cyber-attacks50% of manufacturing companies reported business crime, making it the hardest hit sectorLarger companies are more vulnerable, with 58% of firms employing more than 250 people experiencing crime, compared to 32% of microbusinessesRetail businesses have been particularly affected by shoplifting, with police-recorded incidents rising 20% year on year to reach 516,971 offences in the year to December 2024, exceeding 530,000 by March 2025.Financial Impact on Major CompaniesThe financial consequences of business crime have been substantial, with several high-profile companies suffering significant losses. The hack of Jaguar Land Rover alone is estimated to have cost the UK economy £1.9bn, potentially making it the most costly cyber-attack in British history. Marks & Spencer took a £324m hit to profits after being forced to close its website to orders for more than six weeks following a damaging cyber-attack. Other major companies affected include the Co-op and Booking.com.Industry-Wide Consequences and Economic ImpactCrime against businesses is creating "structural barriers to growth" according to the BCC, forcing companies to divert crucial time and money away from expansion and investment. The impact spans across sectors, from retail and manufacturing to tradespeople experiencing surging tool thefts that threaten their ability to operate. As Ellis Shelton, a policy manager at the BCC, noted, "Bosses are being forced to divert crucial time and money to tackling this anchor on growth."The rising sophistication of criminal activities, particularly in cybercrime and fraud, has left many businesses struggling to keep pace with security measures, especially small and medium-sized enterprises with limited resources.Call for Government Action and Future OutlookIn response to the growing threat, the BCC has called for several specific measures from the government:Creation of a cyber-attack reporting system for companiesEstablishment of regional business crime hubs bringing together police and business crime reduction partnershipsExpansion of cyber and fraud resilience support for small and medium-sized businessesMore incentives for companies to invest in securityWithout decisive action, business crime is likely to continue hampering UK economic growth, with the most sophisticated threats potentially targeting larger companies with greater resources. The BCC's warning suggests that addressing business crime must become a priority for policymakers if the UK is to overcome this "serious barrier" to economic expansion.

The LeadOver 100 new datacentres in the United Kingdom are planning to burn natural gas to generate electricity, with some projects eyeing permanent on‑site generation as a workaround for prolonged grid‑connection delays.The Surge in UK Datacentre Gas RequestsStuart Okin, director of cyber regulation and AI at Ofgem, warned that “there’s 100GW of datacentre projects in the queue” and not all can be linked to the National Grid. Developers therefore “have to come up with an alternative method”.Silvia Simon, head of research at Future Energy Networks, confirmed the firm has received “more than 100” gas‑connection requests in the past two years, many asking for up to 100MW of continuous gas power.Requests total > 15 TWh of energy per year – enough to power London for roughly four and a half months.Projects represent a combined 100GW of planned capacity.The Energy Demand NumbersThe scale of the demand translates into a substantial carbon footprint if supplied by unabated gas. In the United States, similar off‑grid gas generators are projected to emit more CO₂ than the entire nation of Morocco.The Climate and Grid ImplicationsJulian Leslie, director of strategic planning at the UK’s National Energy System Operator (Neso), said the build‑out could jeopardise the Clean Power 2030 goal of keeping unabated gas below 5 % of electricity supply.Eleanor Warburton of Ofgem added that the rapid growth of AI‑driven datacentres is “affecting many aspects of life including energy”, prompting a review of demand‑connection reforms.Environmental groups, such as Action to Protect Rural Scotland (APRS) led by Kat Jones, argue the rush ignores decades of climate science and risks “climate breakdown”.The Path Forward for Policy and AI InfrastructureGovernment and regulators are considering prioritising strategic connections for AI projects while accelerating reforms to speed up viable grid links. If permanent gas generation becomes the norm, further policy measures – possibly including carbon‑pricing or mandatory emissions reporting – may be required to keep the UK on track for its net‑zero commitments.

After a week‑long outage that crippled Canvas for millions of students worldwide, Instructure announced it had reached an agreement with the ransomware group ShinyHunters. While the company stopped short of confirming a payment, the deal raises fresh questions about the wisdom of paying extortionists to protect sensitive educational data. Instructure’s Agreement with ShinyHunters: What Actually Happened The attack began when the group exploited a vulnerability in Instructure’s “Free for Teacher” software, allowing them to deface login pages at institutions such as the University of Texas San Antonio. ShinyHunters threatened to leak 3.6 TB of data – student IDs, emails, names and messages from 9,000 schools and roughly 275 million students and staff – unless a ransom was paid. Instructure later said the stolen data had been “returned” and that it received “digital confirmation of data destruction” via shred logs, but it did not explicitly confirm a payment. Financial Stakes: Ransom Demands, Potential Payments, and Industry Benchmarks ShinyHunters initially demanded $10 million in ransom. Australian ransomware surveys show the average payment fell to $711,000 in 2025, down from $1.35 million the year before. According to a McGrathNicol report, 64 % of surveyed Australian firms had paid a ransom, and 81 % said they would be willing to do so. As of January 2026, 75 Australian businesses with turnovers of at least $3 million had paid ransoms, though the total amount remains undisclosed. Cyber‑security experts estimate that Instructure’s payout – if any – could be anywhere up to the $10 million demand, potentially reduced through negotiation. Policy and Business Implications: Why Paying Ransom Remains Controversial Governments in the UK, US and Australia advise against paying ransoms, arguing that non‑payment reduces the attractiveness of ransomware as a crime vector. In Australia, paying a designated attacker could breach the autonomous cyber‑sanctions law, exposing firms to prosecution on a case‑by‑case basis. Critics also note that payment does not guarantee data will not be leaked; attackers may still copy or sell the information after receiving money. Experts such as Darren Hopkins (McGrathNicol) and Luke Irwin (Aegis Cybersecurity) stress the “trust factor” – criminals must appear honest to receive payment, yet they remain untrustworthy. This paradox fuels boardroom debates about risk‑driven decision‑making versus investing in prevention and incident response capabilities. Looking Ahead: How Companies May Navigate Future Extortion Threats The Canvas case underscores the need for stronger cyber‑resilience strategies: regular vulnerability patching, robust backup architectures, and clear ransomware response playbooks. Insurers are tightening coverage terms, often requiring demonstrable mitigation measures before honoring ransom claims. Policymakers may also tighten reporting obligations and consider clearer prohibitions on ransom payments, especially for critical‑infrastructure providers like education platforms. Ultimately, firms will have to balance the immediate pressure to restore services against the long‑term cost of incentivising criminal enterprises. As ransomware groups refine their extortion tactics, the industry’s collective stance on paying – or refusing – will shape the next wave of cyber‑crime economics.