News & Updates
The latest news and updates from companies in the WLTH portfolio.
White House wants Claude's Mythos model despite Pentagon-Anthropic feud: 'Don't give a f*** about War's position'- Moneycontrol.com
Mythos AI found thousands of major cybersecurity vulnerabilities Claude's powerful new AI model, Mythos Preview, has caught the attention of White House which wants to deploy it across parts of the US federal government even as the Pentagon remains locked in a bitter fued with Dario Amodei's company. Claude had earlier refused to allow its AI to be used for fully autonomous warfare or mass surveillance of Americans, leading to a stand-off with the Trump administration. However, the White House is now engaging with the company to use its new model, which was stopped from being rolled out to public due to its powerful capabilities. Despite tensions, officials acknowledge that the company's technology could bolster national security -- or pose risks if misused, according to a report in Axios. Anthropic is currently limiting access to Mythos Preview to a small group of organisations to study its advanced cyber capabilities and strengthen safeguards. Some US government agencies are now seeking to be part of that group. Mythos has found "thousands" of major vulnerabilities in operating systems, web browsers and other software. Its capabilities to code at a high level have given it a potentially unprecedented ability to identify cybersecurity vulnerabilities and devise ways to exploit them, Reuters quoted experts as saying. According to the Axios report, the Office of Management and Budget has already reached out internally, saying it is reviewing whether agencies can use the model. Two sources told Axios that discussions with Anthropic are underway and access could be granted within weeks. Pentagon stand-off Earlier, the Pentagon barred Anthropic from its contracts, branding it a "supply chain risk" and directing vendors to remove its software from defence workflows. However, the company can continue working with non-military arms of the government while litigation continues. "There's progress with the White House. There's not progress with [the Department of] War," an administration official told Axios. Another official defended the evaluation process but criticised Anthropic's approach. "The government 'has a responsibility to evaluate every model to see where the frontier of tech is,'" the official was quoted as saying by Axios. The official added that the company was using "fear tactics" about Mythos' hacking potential and is "succeeding". At the same time, internal divisions remain stark. "All the intel agencies use Anthropic. Every agency except War (Pentagon) wants to. That's because Anthropic doesn't want to kill people and War's position is 'don't tell us what the f*** to do.' But if you're the Department of Energy, you don't give a f*** about that. You're worried about the Chinese attacking the energy grid. So you want Anthropic," another official told Axios. Anthropic has maintained that its models will not be used for mass surveillance or fully autonomous weapons. Meanwhile, civilian agencies, including those overseeing energy and finance, see potential benefits of Claude's AI Model. Access to Mythos could help identify vulnerabilities in critical infrastructure and improve preparedness against cyber threats. One Defence official previously told Axios that talks continued because "these guys are that good."
White House wants Claude's Mythos model despite Pentagon-Anthropic feud: 'Don't give a f*** about War's position'
Claude's powerful new AI model, Mythos Preview, has caught the attention of White House which wants to deploy it across parts of the US federal government even as the Pentagon remains locked in a bitter fued with Dario Amodei's company. Claude had earlier refused to allow its AI to be used for fully autonomous warfare or mass surveillance of Americans, leading to a stand-off with the Trump administration. However, the White House is now engaging with the company to use its new model, which was stopped from being rolled out to public due to its powerful capabilities. Despite tensions, officials acknowledge that the company's technology could bolster national security -- or pose risks if misused, according to a report in Axios. Anthropic is currently limiting access to Mythos Preview to a small group of organisations to study its advanced cyber capabilities and strengthen safeguards. Some US government agencies are now seeking to be part of that group. Mythos has found "thousands" of major vulnerabilities in operating systems, web browsers and other software. Its capabilities to code at a high level have given it a potentially unprecedented ability to identify cybersecurity vulnerabilities and devise ways to exploit them, Reuters quoted experts as saying. According to the Axios report, the Office of Management and Budget has already reached out internally, saying it is reviewing whether agencies can use the model. Two sources told Axios that discussions with Anthropic are underway and access could be granted within weeks. Pentagon stand-off Earlier, the Pentagon barred Anthropic from its contracts, branding it a "supply chain risk" and directing vendors to remove its software from defence workflows. However, the company can continue working with non-military arms of the government while litigation continues. "There's progress with the White House. There's not progress with [the Department of] War," an administration official told Axios. Another official defended the evaluation process but criticised Anthropic's approach. "The government 'has a responsibility to evaluate every model to see where the frontier of tech is,'" the official was quoted as saying by Axios. The official added that the company was using "fear tactics" about Mythos' hacking potential and is "succeeding". At the same time, internal divisions remain stark. "All the intel agencies use Anthropic. Every agency except War (Pentagon) wants to. That's because Anthropic doesn't want to kill people and War's position is 'don't tell us what the f*** to do.' But if you're the Department of Energy, you don't give a f*** about that. You're worried about the Chinese attacking the energy grid. So you want Anthropic," another official told Axios. Anthropic has maintained that its models will not be used for mass surveillance or fully autonomous weapons. Meanwhile, civilian agencies, including those overseeing energy and finance, see potential benefits of Claude's AI Model. Access to Mythos could help identify vulnerabilities in critical infrastructure and improve preparedness against cyber threats. One Defence official previously told Axios that talks continued because "these guys are that good."
White House wants Claude's Mythos model despite Pentagon-Anthropic feud: 'Don't give a f*** about War's position'
Claude's powerful new AI model, Mythos Preview, has caught the attention of White House which wants to deploy it across parts of the US federal government even as the Pentagon remains locked in a bitter fued with Dario Amodei's company. Claude had earlier refused to allow its AI to be used for fully autonomous warfare or mass surveillance of Americans, leading to a stand-off with the Trump administration. However, the White House is now engaging with the company to use its new model, which was stopped from being rolled out to public due to its powerful capabilities. Despite tensions, officials acknowledge that the company's technology could bolster national security -- or pose risks if misused, according to a report in Axios. Anthropic is currently limiting access to Mythos Preview to a small group of organisations to study its advanced cyber capabilities and strengthen safeguards. Some US government agencies are now seeking to be part of that group. Mythos has found "thousands" of major vulnerabilities in operating systems, web browsers and other software. Its capabilities to code at a high level have given it a potentially unprecedented ability to identify cybersecurity vulnerabilities and devise ways to exploit them, Reuters quoted experts as saying. According to the Axios report, the Office of Management and Budget has already reached out internally, saying it is reviewing whether agencies can use the model. Two sources told Axios that discussions with Anthropic are underway and access could be granted within weeks. Pentagon stand-off Earlier, the Pentagon barred Anthropic from its contracts, branding it a "supply chain risk" and directing vendors to remove its software from defence workflows. However, the company can continue working with non-military arms of the government while litigation continues. "There's progress with the White House. There's not progress with [the Department of] War," an administration official told Axios. Another official defended the evaluation process but criticised Anthropic's approach. "The government 'has a responsibility to evaluate every model to see where the frontier of tech is,'" the official was quoted as saying by Axios. The official added that the company was using "fear tactics" about Mythos' hacking potential and is "succeeding". At the same time, internal divisions remain stark. "All the intel agencies use Anthropic. Every agency except War (Pentagon) wants to. That's because Anthropic doesn't want to kill people and War's position is 'don't tell us what the f*** to do.' But if you're the Department of Energy, you don't give a f*** about that. You're worried about the Chinese attacking the energy grid. So you want Anthropic," another official told Axios. Anthropic has maintained that its models will not be used for mass surveillance or fully autonomous weapons. Meanwhile, civilian agencies, including those overseeing energy and finance, see potential benefits of Claude's AI Model. Access to Mythos could help identify vulnerabilities in critical infrastructure and improve preparedness against cyber threats. One Defence official previously told Axios that talks continued because "these guys are that good."
Perplexity Introduces Personal Computer For Mac -- Here's What The AI Assistant Can Do
Perplexity has launched Personal Computer, which is now available for Mac users. AI company Perplexity has launched a new software called Personal Computer, which is now available for Mac users. This tool builds upon the multi-model orchestration features that the company first introduced with Perplexity Computer at the end of February. Personal Computer includes AI agents capable of interacting with users' files, applications, connectors, and the internet to handle complex tasks and ongoing workflows. What Can Perplexity Personal Computer Do? By bringing the advanced orchestration capabilities of Perplexity Computer directly to users' PCs, Personal Computer can operate across stored files, native apps, external connectors, and web resources. This creates a more personalised experience that combines local and cloud-based environments, improving productivity and security at the same time, Perplexity said. The company highlighted one particularly setup, which involves running it on a Mac mini, where it remains active around the clock. This allows for persistent operations or secure access to private files and native programmes, and users can start or oversee tasks remotely from their phones while on the move. Perplexity has highlighted several applications for the tool. For instance, users can instruct Personal Computer to review their to-do list or even complete the items on it. They can open the Notes app on Mac and request assistance, after which the system determines the best way to help. It can draw from files and work with applications, such as Apple Messages, while deploying multiple agents to finish the job. Users can also ask it to sort disorganised folders by renaming files logically and establishing a clear structure. Interaction with Personal Computer supports voice commands, and users can begin or monitor tasks directly from their smartphones. Perplexity said the app operates within a secure sandbox when creating files, and every action it performs is fully auditable and reversible. Perplexity Personal Computer Availability Personal Computer for Mac is initially rolling out to Max subscribers, with plans to extend availability to other users shortly, giving priority to those who signed up for the waitlist. Also read: Anthropic Launches Claude Opus 4.7 With Improved Benchmark Performance Essential Business Intelligence, Continuous LIVE TV, Sharp Market Insights, Practical Personal Finance Advice and Latest Stories -- On NDTV Profit.
Flight chaos looms over shortage of aviation fuel
Passenger flights could be grounded in the next two months because of a global shortage of aviation fuel caused by the Middle East war, industry experts warned on Thursday. Countries in Asia are likely to be affected first, followed by Europe. Both rely on oil from the Gulf and its refineries for their supplies. "The situation can, within the next three, four weeks, become systemic," energy economist Claudio Galimberti said. "So you can have severe cuts of flights in Europe, already starting in May and June." Flights had already been canceled because of fuel shortages, he said. German airline Lufthansa is closing its regional subsidiary because of higher fuel prices and industrial action. "The 27 operational aircraft of Lufthansa CityLine will be permanently removed from the flight program," the airline said. The Airports Council International Europe has told the European Commission that shortages of jet fuel could start at the beginning of May if tankers did not begin sailing through the Strait of Hormuz before then. Europe has "maybe six weeks or so of jet fuel left," the head of the International Energy Agency said on Thursday. Fatih Birol warned of flight cancelations "soon" if oil supplies remained blocked by the war. It was "the largest energy crisis we have ever faced," he said. "Everybody is going to suffer. Some countries may be richer than the others. Some countries may have more energy than the others, but no country, no country is immune to this crisis. "I can tell you soon we will hear the news that some of the flights from city A to city B might be canceled as a result of lack of jet fuel." The impact of the fuel shortage will vary beween airports and airlines, said another economist, Rico Luman. "Smaller inland airports will be in a weaker position than the main hubs," he said. "It won't be a matter of full halt, but part cancelation at some airlines and airports."
Anthropic releases Claude Opus 4.7: How to try it, benchmarks, safety
Anthropic has been shipping products and making news at a blistering pace in 2026, and on Thursday, the AI company announced the launch of Claude Opus 4.7. Claude Opus 4.7 is Anthropic's most intelligent model available to the general public. Notably, Anthropic said in a press release that Opus 4.7 is not as powerful as Claude Mythos, which Anthropic deemed too dangerous for public release. Claude Opus is a family of hybrid reasoning models capable of multi-step reasoning and advanced coding. Until the announcement of Claude Mythos on April 7, Claude Opus was considered Anthropic's most advanced series of AI models. Don't miss out on our latest stories: Add Mashable as a trusted news source in Google. How to try Claude Opus 4.7 Claude Opus 4.7 is available now via Claude AI, the Claude API, and Anthropic partners such as Microsoft Foundry. The new model is priced the same as Claude Opus 4.6. However, Anthropic noted that because "Opus 4.7 thinks more at higher effort levels," it uses more ouput tokens than its predecessor. Users can read more about how to optimize token usage in the Opus 4.7 migration guide. How Claude Opus 4.7 improves over 4.6 As expected, Claude Opus 4.7 offers improved capabilities across the board. In particular, Anthropic says Claude Opus 4.7 is better at advanced coding tasks, visual intelligence, and document analysis. Anthropic also says Opus 4.7 is "more tasteful and creative when completing professional tasks, producing higher-quality interfaces, slides, and docs." "Users report being able to hand off their hardest coding work -- the kind that previously needed close supervision -- to Opus 4.7 with confidence. Opus 4.7 handles complex, long-running tasks with rigor and consistency, pays precise attention to instructions, and devises ways to verify its own outputs before reporting back," reads an Anthropic blog post. Claude Opus 4.7: Benchmark performance Anthropic released a detailed model card outlining how Claude Opus 4.7 compares to other Anthropic models and frontier models from OpenAI, Google, and xAI. Opus 4.7 lags behind the unreleased Claude Mythos, which Anthropic reports scores significantly higher on common benchmarks such as Humanity's Last Exam. "Claude Opus 4.7 is less capable than Claude Mythos Preview on every relevant axis we measured and does not advance our capability frontier," the model card states." That means Claude Opus 4.7 is not evidence that AI development has accelerated beyond existing trend lines. On Humanity's Last Exam (without tools), Anthropic reports that Claude Opus 4.7 outperforms all other frontier models except Claude Mythos. Claude Mythos scored 56.8 percent on HLEClaude Opus 4.7 scored 46.9 percentGemini 3.1 Pro scored 44.4 percentGPT-5-4 Pro scored 42.7 percentClaude Opus 4.6 scored 40.0 percent With tools, GPT-5-4-Pro scored 58.7 percent compared to Opus 4.7's 54.7 percent. Mythos beat them both with 64.7 percent. Mashable has not independently verified these benchmark results. Full results are available in the Opus 4.7 model card. Overall, Anthropic scored Opus 4.7 above other leading models in some benchmarks, though Gemini 3.1 Pro and GPT-5-4 score higher in some areas. Claude Opus 4.7: Safety and hallucinations Anthropic also reports that Opus 4.7 shows a low risk of misaligned behaviors, with a similar risk profile as Opus 4.6. For example, Anthropic says Opus 4.7 is less likely to hallucinate and shows lower rates of reward hacking. "Claude Opus 4.7 is more reliably honest than Opus 4.6 or Sonnet 4.6, with large reductions in the rate of important omissions, and moderate improvements in factuality and rates of hallucinated input," the model card states. Want to learn more about getting the best out of your tech? Sign up for Mashable's Top Stories and Deals newsletters today.
Musk's SpaceX to launch Rosalind Franklin rover to Mars aboard Falcon Heavy in 2028
Elon Musk's SpaceX will launch the Rosalind Franklin to Mars in late 2028. The rover will hitch a ride to the Red Planet aboard the company's Falcon Heavy rocket. Meanwhile, Nasa has cleared a major milestone for one of the most ambitious Mars missions of the decade, approving the implementation phase of the Rosalind Franklin Support and Augmentation (ROSA) project. The move strengthens collaboration between Nasa and the European Space Agency on the long-awaited Rosalind Franklin rover, which had been derailed due to the Russia-Ukraine war. The mission, led by ESA, represents a significant step forward in the search for life beyond Earth. Unlike previous Mars rovers, Rosalind Franklin will drill beneath the planet's surface, up to two meters deep, to look for signs of past or present microbial life. Scientists believe that subsurface samples may hold preserved organic material shielded from harsh radiation on the Martian surface. Under the ROSA project, Nasa will provide critical support elements to ensure mission success. These include launch services, braking engines for the lander platform, and radioisotope heater units that will help keep the rover's internal systems warm in Mars' extreme cold. Nasa will also contribute advanced electronics and a sophisticated mass spectrometer, a key component of the Mars Organic Molecule Analyzer instrument. This instrument will analyze soil samples collected at the rover's landing site, Oxia Planum, a region believed to have once contained water and clay-rich deposits. The partnership between Nasa and ESA was formalized through a Memorandum of Understanding signed in early 2024, expanding Nasa's role in the ExoMars program after earlier mission delays. The ROSA project has since cleared key technical reviews, including its Preliminary Design Review, signaling that development is on track. Nasa has selected SpaceX to provide launch services using Falcon Heavy, one of the most powerful operational rockets in the world. The launch will take place from Launch Complex 39A at Kennedy Space Center, a historic site that has supported missions from Apollo to modern deep-space exploration. The contract for the mission was awarded under the agency's Launch Services II framework. The launch window is currently targeted for no earlier than late 2028. With its unique ability to probe beneath the Martian surface, the Rosalind Franklin rover could redefine humanity's understanding of life beyond Earth, marking a new chapter in international space exploration.
Anthropic Unveils Claude Opus 4.7: Try It, Safety Features, and Benchmarks
In a significant move, AI company Anthropic has launched Claude Opus 4.7, their latest model designed for general use. This new iteration aims to enhance user experience and tackle advanced tasks while balancing safety and reliability. Claude Opus 4.7 Overview Released in 2026, Claude Opus 4.7 is positioned as Anthropic's most intelligent model available. However, it is noted that it does not surpass the capabilities of Claude Mythos. This model has been avoided for public release due to safety concerns. Key Features of Claude Opus 4.7 * Hybrid reasoning models with multi-step reasoning and advanced coding capabilities. * Improved in advanced coding tasks, visual intelligence, and document analysis. * Higher quality in professional outputs like interfaces, slides, and documents. How to Access Claude Opus 4.7 Users can access Claude Opus 4.7 through various platforms, including: * Claude AI * Claude API * Anthropic partners like Microsoft Foundry The model is priced similarly to its predecessor, Claude Opus 4.6. Due to its advanced processing capabilities, it utilizes more output tokens, which users should consider when managing their token limits. Benchmark Performance Benchmark testing has shown that Claude Opus 4.7 performs well compared to other models but falls short of the unreleased Claude Mythos. The following performance metrics were reported: When additional tools were used, GPT-5-4 Pro scored 58.7%, while Opus 4.7 achieved 54.7%. Mythos notably outperformed both with a score of 64.7%. Safety Features and Improvements Anthropic emphasizes that Claude Opus 4.7 maintains a low risk of misalignment, akin to its predecessor. It is less likely to hallucinate and shows reductions in reward hacking incidents. The model card highlights: * Increased reliability and honesty. * Significant reductions in important omissions. * Notable improvements in factual accuracy. Overall, Anthropic's Claude Opus 4.7 marks a significant step in AI development, offering users advanced functionalities while ensuring safety and reliability.
OpenAI signs $20B deal with chip start-up Cerebras
In January, OpenAI had agreed to buy up to 750 megawatts (MW) of computing capacity from Cerebras over three years. The deal was worth more than $10 billion. However, the latest commitments exceed this earlier agreement with the chipmaker. As part of the deal, OpenAI will get warrants for a minority stake in Cerebras. Its ownership could increase as its spending rises. The report also said that OpenAI has agreed to give Cerebras about $1 billion to help fund the development of data centers that would run its AI products. Advertisement The deal between OpenAI and Cerebras underscores a growing demand for computing power to run inference, the process by which AI models generate responses. This comes as companies race to develop reasoning models and applications aimed at driving wider adoption of AI technology.
OpenAI signs massive AI chip deal with Cerebras | News.az
OpenAI is set to spend more than $20 billion on AI chips from startup Cerebras Systems over the next three years, in a deal that could also give the ChatGPT maker a significant ownership stake in the company, according to a report by The Information. The agreement highlights the escalating race for computing power in artificial intelligence, as companies invest heavily in infrastructure to train and run increasingly complex models, News.Az reports, citing Reuters. Under the reported deal, OpenAI will purchase servers powered by Cerebras chips, potentially doubling its previous commitment with the chipmaker. Earlier this year, OpenAI agreed to buy up to 750 megawatts of computing capacity from Cerebras in a deal valued at over $10 billion. The new agreement could push OpenAI's total spending with Cerebras to as much as $30 billion over three years. As part of the arrangement, OpenAI is expected to receive warrants that could translate into a minority equity stake in Cerebras -- potentially up to 10% ownership depending on how much it ultimately spends. The report also said OpenAI could provide around $1 billion to help fund the development of new data centres that will run its AI systems. Neither OpenAI nor Cerebras commented on the report. The deal underscores the growing importance of "inference" computing -- the process by which AI models generate responses -- as demand surges for tools like chatbots, copilots and generative AI services. Cerebras, known for its wafer-scale chip technology, is positioning itself as a challenger to industry leader Nvidia in the high-performance AI hardware space. The partnership is seen as crucial to Cerebras' plans to go public. The company is reportedly preparing for an IPO as soon as the second quarter and is also seeking to raise about $3 billion at a valuation of roughly $35 billion. Founded in 2015 and based in California, Cerebras has attracted attention for building some of the largest and most powerful AI chips in the industry. The potential tie-up with OpenAI signals how leading AI developers are increasingly locking in long-term access to specialised hardware as competition intensifies across the sector.
Europe has just 6 weeks of jet fuel left, warns IEA - What passengers need to know ahead of summer flight chaos
Europe could be heading toward a serious aviation fuel crunch, with the head of the International Energy Agency warning that the region has "maybe six weeks of jet fuel left" if supply problems continue. The warning comes as the Strait of Hormuz, one of the world's most important shipping routes for oil and jet fuel, has remained effectively closed by Iran for more than six weeks. Tehran shut the route in response to US and Israeli attacks, sending fuel prices sharply higher and raising fears of shortages across Europe. Why Europe is worried The International Energy Agency (IEA), which advises 32 countries on energy supply and security, said in its latest monthly report that Europe's fuel stocks could reach a breaking point by June. That could happen if European countries fail to replace at least half of the jet fuel they usually import from the Middle East. Fatih Birol, the executive director of the IEA, told AP that flight cancellations could soon become a real possibility if supplies remain blocked. Europe depends heavily on Middle East fuel The IEA said Europe has historically relied on the Middle East for around 75% of its jet fuel imports. The Gulf region is one of the biggest suppliers of aviation fuel to the global market, making the Strait of Hormuz especially important. The problem does not end there. Refineries in other exporting countries such as South Korea, India and China also depend heavily on crude oil from the Middle East. Because of that, the IEA said the crisis "has thrown a proverbial wrench into the inner workings of the aviation fuel markets". What could happen next? If Europe cannot replace more than 50% of its Middle East imports, "physical shortages may emerge at select airports, resulting in flight cancellations, and demand destruction." Even if Europe manages to replace three-quarters of lost supplies, similar problems could still happen by August. The agency said European markets now need to work harder to attract extra cargoes from elsewhere if enough fuel is to be available through the busy summer season. Many airlines are already struggling with higher fuel costs, which usually make up 20% to 40% of their operating expenses. The benchmark European jet fuel price hit a record high of $1,838 per tonne at the start of April. Before the war began, it stood at $831 per tonne. Flight cancellations could come soon Birol warned that if the route stays blocked, cancellations may begin "soon". European countries are now rushing to find fuel from other sources. Analysts say replacement supplies are mainly coming from the United States and Nigeria. The IEA noted that US jet fuel exports have risen quickly in recent weeks. Still, the agency warned that even if all those US shipments were sent to Europe, they would replace only a little more than half of the lost Gulf supplies. But even with rising US exports, the IEA said those shipments would replace only a little more than half of the missing Middle Eastern supplies. Airlines have already started reacting KLM said it will cancel 160 flights, or 80 return services, across Europe in May because of rising fuel costs linked to Middle East tensions. The airline said this is less than 1% of its European schedule and stressed it is not facing a physical fuel shortage. Lufthansa is speeding up the retirement of 27 older aircraft from its CityLine regional unit. The carrier said the move is linked to doubled kerosene prices and labour strikes. It is part of a bigger plan that could ground 2.5% to 5% of its fleet if the crisis worsens. Lufthansa said grounding more aircraft "may be unavoidable." Other airlines, including Ryanair and Scandinavian Airlines, have also cut hundreds of flights in recent weeks. Some regional airlines have added fuel surcharges. What Europe and the UK are saying The European Commission said there is "no evidence of fuel shortages" in the EU right now. It added that crude oil supplies to refineries remain stable, but admitted there could be supply problems "in the near future." Officials said energy coordination groups are meeting every week, with new measures expected next week. The UK government said it is working with suppliers and airlines "to ensure people keep moving and businesses are supported." UK airlines said there is currently no disruption to fuel supply. What travellers should know now * Passengers are being advised to keep checking flight status, as schedules may change quickly. So far, cancellations remain limited and are mostly linked to cost rather than a lack of fuel. But that could change in May and June, especially at smaller airports. * Travellers may also face higher fares and fuel surcharges. EasyJet alone reported £25 million in extra fuel costs for March, despite having fuel prices locked in earlier. * Major hubs such as Heathrow are expected to be safer because they would likely get fuel first. * If a flight is cancelled, EU261 and UK261 passenger rules normally apply. However, airlines are trying to have fuel-related cancellations treated as extraordinary circumstances, which could reduce compensation rights. * Travel insurance that covers disruption may be useful. Summer holidays could be affected The IEA and analysts say the pressure may grow just as Europe enters the busiest holiday travel season. Flexible tickets, backup plans, and staying updated through airline apps may help passengers manage any sudden changes.
Claude Opus 4.7 is Anthropic's most powerful public model yet
Anthropic has launched its latest artificial intelligence (AI) model, the Claude Opus 4.7. The new version is said to be an improvement over its predecessors. It is better at software engineering, following instructions, and completing real-world work. Claude Opus 4.7 is also the most powerful model available to the public from Anthropic. However, its cyber capabilities aren't as advanced as those of Claude Mythos Preview, which was released earlier this month to select companies under Project Glasswing, Anthropic's cybersecurity initiative.
Scott Wedgewood stops 22 shots, Avs beat Kraken 2-0 to break franchise's single-season points mark
DENVER (AP) -- Nick Blankenburg scored a second-period goal, Scott Wedgewood made 22 saves for his fourth shutout of the season and the Colorado Avalanche beat the Seattle Kraken 2-0 on Thursday night to break the franchise's single-season points record. The Avalanche won the Presidents' Trophy with 121 points, eclipsing the total of 119 points set by the 2021-22 squad that went on to win the Stanley Cup.
Anthropic Unveils Claude Opus 4.7 for Public Use, Expands AI Capabilities
Anthropic launches Claude Opus 4.7, bringing stronger coding, visual reasoning, and safeguards while making advanced AI accessible to everyday users. Anthropic has introduced its latest artificial intelligence model, Claude Opus 4.7, marking a significant upgrade in its Opus series and making advanced AI capabilities more accessible to general users. The launch comes shortly after the company revealed its highly advanced Claude Mythos model, which it has restricted to a limited group of organizations due to its powerful capabilities. Claude Opus 4.7 represents a notable improvement over its predecessor, Opus 4.6, particularly in areas such as coding, instruction-following, and visual reasoning. While it is not as broadly capable as the Mythos model, Anthropic positions Opus 4.7 as a strong middle-ground solution -- offering enhanced performance without the risks associated with more powerful systems. The model is designed to handle complex and long-running tasks with greater precision and consistency. This makes it especially useful for software engineering and real-world problem-solving, where earlier versions required more human oversight. With improved instruction adherence, Opus 4.7 can better interpret and execute user prompts, reducing ambiguity and increasing reliability. One of the key upgrades in Opus 4.7 is its enhanced visual processing capability. The model can now analyze high-resolution images up to 2,576 pixels, allowing it to extract detailed information from dense screenshots, diagrams, and other visual data formats. This marks a significant step forward compared to previous versions, which often struggled with complex visuals. The release also brings improvements to Claude Code, a widely used tool among developers. Opus 4.7 enables more autonomous task execution, potentially reducing the need for continuous human supervision during coding workflows. However, this advancement comes with increased computational demands. The model uses more processing tokens, prompting Anthropic to raise usage limits for subscribers. Boris Cherny, the head of Claude Code, addressed this change on X, stating, "Opus 4.7 uses more thinking tokens, so we've increased rate limits for all subscribers to make up for it. Enjoy!" Despite its advancements, Opus 4.7 is distinct from the Mythos model. Anthropic clarified that it is not derived from Mythos but instead serves as a direct successor to Opus 4.6. While Mythos focuses heavily on cybersecurity capabilities, Opus 4.7 incorporates stricter safeguards to prevent misuse. The model is designed to automatically block "high risk" cybersecurity-related requests, reducing the likelihood of exploitation by malicious users. The company's cautious approach to Mythos stems from concerns about its potential to identify vulnerabilities at an advanced level, which could be misused if widely available. As a result, Mythos remains limited to around 40 select organizations under Project Glasswing. Claude Opus 4.7 is now available across Anthropic's ecosystem, including its native Claude platform and major cloud services such as Amazon Bedrock, Google Cloud's Vertex AI, and Microsoft Foundry. Notably, the pricing remains unchanged from the previous version, set at $5 per million input tokens and $25 per million output tokens. The announcement has already had ripple effects in the market, with companies like Adobe and Figma reportedly seeing stock declines amid expectations that the new model could disrupt design workflows by automating complex creative tasks.
OpenAI Supercharges Codex Into 'AI Super App', Sam Altman Takes Swipe at Anthropic Amid Tech Rivalry
OpenAI is transforming Codex into a powerful AI super app by merging ChatGPT, coding tools, and browser features. CEO Sam Altman also takes a subtle dig at rival Anthropic as the AI battle intensifies. OpenAI has just turned its coding assistant Codex into something far bigger than a developer's tool. With its latest update, Codex now behaves more like a super app for work automation, capable of running desktop applications, browsing the web, generating images, and even scheduling tasks- all without constant human supervision. The update introduces background computer control on macOS, meaning Codex can autonomously click, type, and operate apps. It's no longer limited to writing snippets of code; it can now design front-end layouts, create gaming assets, and even draft product concepts. OpenAI has also added an in-app browser and support for image generation via gpt-image-1.5, giving Codex a creative edge that Anthropic's competing tool, Claude Code, currently lacks. Another major leap is automation. Codex can schedule future work and wake itself up to continue long-term tasks, effectively acting like a digital colleague who doesn't need breaks. With over 90 integrated plugins, from Atlassian to Microsoft Suite, the app is positioned as a full-fledged work agent rather than just a coding assistant. The timing of the release was notable. It came just hours after Anthropic unveiled Claude Opus 4.7, its latest AI model. OpenAI CEO Sam Altman wasted no time in making his presence felt. Quoting Codex chief Thibault "Tibo" Sottiaux on X, Altman joked about not wanting to face reduced limits or weaker models. While he didn't name Anthropic directly, the remark was widely read as a jab at rival CEO Dario Amodei, whose company has faced criticism for restricting Claude Code users and nudging them toward less powerful models. This isn't the first time the two leaders have clashed. Earlier this year, Altman and Amodei famously refused to share a stage at the India AI Impact Summit, underscoring the competitive tension between OpenAI and Anthropic. Codex's new features put it head-to-head with Claude Code in the race to dominate automated workflows. Anthropic's tool has been praised for its coding ability but lacks image generation and deep automation. OpenAI's update directly addresses those gaps, positioning Codex as a more versatile platform. OpenAI is clear about its ambitions: Codex isn't just for developers anymore. The company wants it to be the go-to work agent, capable of handling everything from software engineering to creative design. With more than 3 million active users already, the expansion could cement Codex as one of the most widely adopted AI productivity tools. The rollout is now live for Codex desktop app users signed in with ChatGPT, marking a significant step in OpenAI's strategy to blend coding, creativity, and automation into one seamless experience.
OpenAI to spend more than USD 20 billion on Cerebras chips, receive stake - The Tribune
OpenAI has agreed to pay chip startup Cerebras more than $20 billion over the next three years to use servers powered by the company's chips, under a deal that could also give the ChatGPT maker an equity stake in the firm, The Information reported on Thursday, citing sources. The development comes as OpenAI attempts to pull ahead in the AI race and meet growing demand. In January, the company agreed to buy up to 750 megawatts of computing capacity from Cerebras over three years in a deal valued at more than $10 billion. The newer commitments are double the size of OpenAI's previously reported agreement with the chipmaker. Reuters could not independently verify the report. OpenAI did not respond to a request for comment outside regular business hours, while Cerebras declined to comment. The deal highlights the industry's growing appetite for computing power to run inference - the process by which AI models generate responses. Cerebras could disclose parts of its previously undisclosed arrangement with OpenAI as soon as Friday, the report said. Under the deal, OpenAI will receive warrants for a minority stake in Cerebras, with its ownership potentially increasing as its spending rises, The Information reported. It added that OpenAI has also agreed to provide Cerebras about $1 billion to help fund the development of data centers that would run its AI products. The company's total spending over the next three years could reach $30 billion, which may translate into warrants representing up to a 10% stake in Cerebras, the report added. Cerebras' reliance on deal, IPO plans The tie-up with OpenAI is central to Cerebras efforts to go public, with the AI chipmaker targeting a listing in the second quarter of this year. Sunnyvale, California-based Cerebras, last valued at $23.1 billion, also plans to raise $3 billion in an offering next month at a valuation of about $35 billion, The Information reported on Thursday. Founded in 2015, the company is known for its wafer-scale engine chips and competes with products from Nvidia and other AI chipmakers. OpenAI CEO Sam Altman is an early investor in Cerebras.
US Agencies Eye Anthropic's Mythos AI Despite Trump's Ban
The US government is quietly preparing to roll out Anthropic's powerful frontier AI model, Mythos, across federal agencies even though the Trump administration previously cut ties with the company. Anthropic unveiled Mythos earlier this month under Project Glasswing, a controlled program granting select organizations access to the unreleased Claude Mythos Preview model. The AI has already proven its muscle, reportedly uncovering thousands of vulnerabilities in operating systems, browsers, and widely used software. Its ability to write and analyse code at a high level makes it both a defensive asset and a potential risk if misused. Gregory Barbaccia, federal chief information officer at the White House Office of Management and Budget (OMB), confirmed in an internal email that safeguards are being set up to allow agencies to experiment with Mythos. He emphasised that the administration is working closely with model providers, industry partners, and the intelligence community to ensure "appropriate guardrails" before releasing a modified version of the tool. Officials have stressed that any new AI technology must undergo rigorous evaluation for fidelity and security before being integrated into government systems. Anthropic's relationship with the federal government has been rocky. Co-founder Jack Clark revealed that the company had discussions with the Trump administration about Mythos, even after the Pentagon severed business ties following a contract dispute. That earlier rupture effectively banned Anthropic from Pentagon work, but the current push to deploy Mythos suggests agencies are willing to sidestep Trump's ban in favor of addressing urgent cybersecurity needs. The decision highlights a broader tension: balancing political decisions of the past with the pressing need to secure America's digital infrastructure today. Mythos's ability to expose weaknesses in critical systems makes it a double-edged sword capable of strengthening defenses but also dangerous if exploited. With Reuters inputs
Musk's SpaceX is shaping up as the biggest IPO on record. It's also bending the rules to do so
Add Yahoo as a preferred source to see more of our stories on Google. Elon Musk's space exploration company SpaceX has filed confidential papers ahead of a planned public company listing on the US NASDAQ stock exchange. The initial public offering (IPO) for the company controlled by the world's richest man is targeting a total valuation of US$2 trillion. Musk plans to list only a small fraction of the company to raise US$75 billion from public investors, which would still make it the largest IPO in history. So, why is SpaceX planning to go public? And what does the IPO mean for investors who might want a tiny slice of the action? SpaceX says it aims to "make humanity multiplanetary". You would expect no less from Musk, who founded SpaceX in 2002. His company's breakthrough was to re-use as much of the rocket and launcher vehicle as possible. This slashed launch costs to as little as 5% of the costs in the early 2000s, and turned commercial space flight from science fiction into reality. The company says it has now completed about 600 successful rocket landings. Yet, for all its space ambitions, SpaceX still derives 50-80% of its revenue from Starlink, a communications business, which provides satellite internet to over 10 million users around the world. In February 2026, SpaceX merged with xAI, the loss-making AI company behind the Grok chatbot, in what was the largest private merger transaction on record. The deal valued xAI at US$250 billion and SpaceX at US$1 trillion, creating a combined entity worth US$1.25 trillion. The merger has helped to set the stage for the SpaceX IPO. Musk suggested the IPO proceeds will be used for launching up to one million data centre satellites into space. The idea is that space-based data centres would be powered by abundant solar energy, and therefore bypass the constraints of electricity and water usage on Earth. SpaceX may be the first of three mega-IPOs this year, ahead of potential listings of AI companies Anthropic and OpenAI. If it goes ahead with plans to raise US$75 billion, that would represent just 3.75% of the company's total value. It means the vast majority of SpaceX would remain in private hands, owned by Musk himself and a handful of early private investors. In stock market terms, this is called a low "free float". Normally, companies that only list such a small percentage of their total value would not qualify for inclusion in major stock market indices like the S&P 500 or the NASDAQ 100. The NASDAQ normally requires at least a 10% free float of shares in a given company. But to allow a potential listing of SpaceX to be included in the index, the exchange has introduced a special adjustment to the weighting of shares and removed the 10% minimum. NASDAQ also reduced the normal "seasoning period" before a newly listed company can join the index from three months to just 15 trading days. Again, this is to accommodate the SpaceX listing. For investors in passive funds, including exchange-trade funds (ETFs), this matters a lot. Currently, more than US$600 billion of investors' money is with passive funds that track the NASDAQ 100 index. As soon as SpaceX joins the index, these investors will automatically be buying in. The concern is that allowing giant companies such as SpaceX to enter the index too quickly could lead to big price swings, which would expose millions of investors to high volatility. SpaceX wants investors to value it at US$2 trillion, but it only earned US$15 billion in revenue last year. At that rate, it would take 133 years of revenue just to match its current asking price. Tesla, one of the most expensive stocks in the world, would take just 13 years -- making SpaceX's price tag ten times higher. Other leading market indices, such as S&P 500 and FTSE Russell, are also bending their rules to fast-track the inclusion of very large, newly listed companies. Many more investors have their money in funds that track S&P indices compared to Nasdaq 100 - more than US$16 trillion in passive funds track the S&P. If the S&P 500 follows NASDAQ's lead and changes its own rules to accommodate SpaceX, the wave of automatic buying would be even larger. Musk's companies have long been the darlings of non-professional, retail investors, and SpaceX would be no exception. In fact, the company said it aims to sell up to 30% of its shares to non-institutional, individual investors. With SpaceX's sky-high valuation, investors need to stop and think before buying in. But when powerful companies can rewrite the rules in their own favour, thinking carefully becomes a luxury. Markets only work when everyone plays by the same rules, and right now, not everyone is.
Anthropic's Mythos ups the stakes for IT cos, signals deeper disruption
Bengaluru: Anthropic's Mythos marks a clear break from the incremental gains typically seen across successive frontier AI systems, particularly in software engineering tasks, with significant implications for the sector.In a recent note, Kotak Securities said that if these improvements translate effectively into real-world enterprise deployments, its earlier estimate of a 3-3.5% annual growth headwind for the IT services industry over the next three years could shift from a cautious projection to a more realistic baseline. It also warned of rising downside risks if future frontier models continue to deliver step-change improvements in capability.The note highlighted that Mythos demonstrates a sharp jump in benchmark performance across software engineering tasks, diverging from the recent trajectory of moderate, incremental progress. Based on qualitative assessments, the model appears to significantly advance agentic software development capabilities. However, Kotak cautioned that the model's real-world impact remains uncertain, as its capabilities are yet to be proven at scale due to the absence of a public release. It added that disruption risks could be higher for IT services firms with greater exposure to application services. Industry voices suggest the impact could extend beyond traditional IT services. Viral Shah, co-founder and CEO of JuliaHub, said even engineering services firms focused on hardware design, testing and verification of aircraft, automobiles and semiconductors could face disruption. "Anything that is digital is a fair game. AI represents a shift from labour to capital," he said. While the short-term impact may be seen in margin pressures, the long-term consequences could be existential. "I expect a bimodal distribution where a few firms adapt and go from strength to strength while others may vanish," he added. Shah also described Mythos as a major advancement, noting that it has identified security vulnerabilities in highly secure operating systems that had previously gone undetected.At the same time, some experts see a shift in demand rather than an outright decline. Namratha Dharshan, chief business leader at ISG, said revenue deflation has already occurred in routine, repetitive and low-value work. However, she noted that AI-driven disruption is also creating new opportunities in areas such as integrations, data readiness, governance, cybersecurity and workflow design. ISG is already seeing a 20% year-on-year increase in new scope of work as companies pivot to these segments. "Labour arbitrage is not a sustainable model going forward," she said.Analysts also point to a structural shift in service delivery. Biswajeet Mahapatra, principal analyst at Forrester, said services built around repetitive defect discovery, regression testing and low-complexity application maintenance are the most exposed. Mythos-class models can significantly compress the effort required to identify bugs, security flaws and logic errors, particularly in code-heavy environments.However, he does not expect a wholesale revenue collapse. Instead, the bottleneck is likely to shift from detection to areas such as triage, remediation prioritisation, coordinated patching and safe deployment -- functions that still require deep system context, governance and accountability. Firms that remain dependent on manual testing models may face risks, while those that pivot to secure-by-design engineering, remediation orchestration and AI-augmented reliability services are likely to see demand evolve rather than disappear.
OpenAI targets Anthropic with Codex upgrade that expands desktop control
OpenAI is stepping up its fight with Anthropic in the race to dominate AI coding tools, rolling out a significantly upgraded version of Codex that pushes it closer to being an autonomous workplace assistant. The move comes as Anthropic's Claude Code continues to gain traction among businesses, widely seen as the current frontrunner, according to a TechCrunch report.The most notable shift is how Codex operates. It can now run in the background on a user's computer, opening apps and performing actions with a cursor that clicks and types -- effectively behaving like a digital co-worker.This enables multiple AI agents to run simultaneously, working in parallel without disrupting the user's own tasks. OpenAI positions this as a "coding buddy" that handles auxiliary work -- iterating on front-end changes, testing applications, or navigating tools that lack APIs -- while developers focus on higher-level problems.The update also brings Codex closer to capabilities already showcased by Anthropic. Tools like Claude Code have demonstrated remote desktop control, allowing AI to act independently even when users are away.OpenAI's additions suggest a clear response: not just keeping pace, but embedding Codex deeper into day-to-day workflows. The new in-app browser, for instance, lets users issue commands that the agent executes across web apps, with plans to expand beyond local environments.A key addition is a "memory" feature that allows Codex to recall previous sessions and adapt to how a user works over time. This signals a shift from reactive assistance to personalised, context-aware automation.OpenAI is also expanding Codex's utility through 111 plug-in integrations, including tools like CodeRabbit and GitLab Issues. The idea is straightforward: let the AI handle routine coordination. As TechCrunch notes, Codex could soon scan Slack messages and Google Calendar events to generate a daily to-do list -- essentially taking over light clerical work.The tool has also gained image-generation capabilities for creating mockups, slide visuals, and product concepts, extending its reach beyond pure coding tasks.Alongside these features, OpenAI has introduced a pay-as-you-go pricing model for enterprise and business users, signalling a push towards flexibility and broader adoption.Once dominant in consumer AI, OpenAI is increasingly prioritising enterprise use cases, even as it scales back projects like its social video experiment, Sora 2. The company is also navigating legal and reputational challenges tied to ChatGPT's real-world impact.
