Hyundai Creta Summer Edition: Hyundai has introduced the Summer Edition of one of its highest-selling SUVs, the Creta. It is offered with additional features across select trims of the midsize SUV. Hyundai Creta Summer Edition: Hyundai has introduced the Summer Edition of one of its highest-selling
The Punjab and Haryana High Court has granted bail to a man accused of passing on sensitive information about Operation Sindoor to Pakistan. Justice Vinod S Bharadwaj noted on April 1 that the prosecution had failed to submit any material to substantiate its allegations against Davender Singh
The first thing you notice at sea is the rhythm, not just the movement of the waves, but the quiet way life begins to fall into sync with them. Suddenly, mornings feel slower; evenings stretch longer, and somewhere between the horizon and the sky
As temperatures rise, there’s nothing quite like a cold beer to cut through the heat. From bold strong brews to crisp, easy-drinking lagers and fruit-forward picks, this season’s beer lineup is all about versatility and refreshment. Whether you’re planning laid-back evenings, match screenings
Project Glasswing aims to secure the world’s software, and Anthropic’s mythologyAnthropic is pitching Project Glasswing as a cybersecurity breakthrough, but the first battle may be getting everyone to believe in the legend of Claude Mythos Published on: Apr 09
West Bengal election 2026: Shah’s extended stay in West Bengal points to a detail-oriented campaign approach. The focus is likely to be twofold: consolidating strongholds where the BJP already has a firm base, and aggressively targeting constituencies it narrowly lost in 2021
Jaya Bachchan is celebrating her 78th birthday today. She married legendary actor Amitabh Bachchan on 3 June 1973. New Delhi: Actor-turned-politician Jaya Bachchan has turned 78 today. She has been married to superstar Amitabh Bachchan for over five decades now
Ashish Nehra’s explosive reaction after Shubman Gill’s overthrow goes viral as GT beat DC in a last-ball IPL 2026 thriller. The moment that’s dominating IPL chatter right now is Ashish Nehra’s explosive on-camera reaction after Shubman Gill’s run-out attempt turned into four overthrows
Project Glasswing aims to secure the world’s software
Project Glasswing aims to secure the world’s software, and Anthropic’s mythology
Project Glasswing aims to secure the world’s software, and Anthropic’s mythology
Anthropic is pitching Project Glasswing as a cybersecurity breakthrough, but the first battle may be getting everyone to believe in the legend of Claude Mythos
Published on: Apr 09, 2026 1:14 PM IST By Vishal Mathur Share via Copy link There’s an irony underlying this situation. An artificial intelligence (AI) model that Anthropic was working on was ‘leaked’ a few days ago. Codenamed Mythos, and capabilities claimed to be “substantially beyond those of any model we have previously trained, this model was discovered after descriptions of the model were found in publicly accessible data following an ‘error’ in the AI company’s content management system. Now, Anthropic says Claude Mythos is so powerful, they’ll not make this model generally available. Instead, it provides the foundation for something called Project Glasswing, to secure every other software. Anthropic says they’ll be committing up to $100 million worth of usage credits for the Mythos Preview across these efforts. (Representative file photo) Where have you heard this before? In 2019, Dario Amodei who is now the CEO and co-founder of Anthropic was still at OpenAI as Vice President of Research. In February 2019, OpenAI had said that its text-generation GPT-2 algorithm was too dangerous for public release. Has the marketing playbook been carried forward with the passage of time? now, this model is the basis for an industry consortium, called Project Glasswing, with an intention to deploy this for cybersecurity “Instead, we are using it as part of a defensive cybersecurity program with a limited set of partners,” the company details, in the system card for the Claude Mythos Preview released this week. The partners include Apple, Google, Microsoft, Amazon Web Services, Nvidia and Cisco. Anthropic insists the model’s “striking leap in scores on many evaluation benchmarks” is the reason for the advantage over Claude Opus 4.6 across agentic coding, reasoning and computer use benchmarks. The GPT-2 resemblance seems difficult to ignore. Back then, the rhetoric was more about AI responsibility and restraint. Eventually, GPT-2 was released (small model in Feb 2019, the medium model in May and the larger models in August and November) , relegating the ‘too dangerous to share’ pitch more to the AI marketing folklore, than a pure safety debate. The Project Glasswing revives those memories, as a model that’s claimed to be exceptionally capable and therefore very risky, with a need for controlled usage. “As part of Project Glasswing, the launch partners listed above will use Mythos Preview as part of their defensive security work; Anthropic will share what we learn so the whole industry can benefit. We have also extended access to a group of over 40 additional organisations that build or maintain critical software infrastructure so they can use the model to scan and secure both first-party and open-source systems,” the company says in a statement. Anthropic says they’ll be committing up to $100 million worth of usage credits for the Mythos Preview across these efforts—after that, Claude Mythos Preview will be available to participants at $25/$125 per million input/output tokens. Important to note the commercial architecture that’s being created here, with Claude Mythos still very much a product in Anthropic’s scheme of things—scarcity, capabilities, and therefore premium pricing strategies will be at work. They are bullish about their claims, saying the model found a 27-year-old vulnerability in OpenBSD operating system and a 16-year-old vulnerability in FFmpeg software used to encode and decode video. They also say that the Mythos Preview has already found thousands of high-severity vulnerabilities—including some in every major operating system and web browser. The partners insist this will be one part of their overall cybersecurity efforts. “We will take a rigorous, independent approach to determining how to proceed and where we can help,” says Pat Opet, Chief Information Security Officer at JPMorganChase. Heather Adkins, VP of Security Engineering at Google, says they’ll make the Mythos Preview available via the Vertex AI platform and “will continue investing in our leading cybersecurity platform and a culture focused on protecting users.” Amy Herzog, Amazon Web Services’ Vice President of CSIO, notes that their teams analyse over 400 trillion network flows every day for threats, and AI is crucial for identifying and defending at scale. They’ve applied Claude Mythos Preview to these security operations, and it has helped strengthen the code in critical instances. The unwritten caveat must be noticed. None of the Project Glasswing partners are fully handing over their cybersecurity apparatus or architecture to the new model. This will be an additional layer, not the entirety. This lends the necessary perspective. Claude Mythos’ claimed scores and capabilities are impressive, but it is unlikely that one AI model will fundamentally restructure the approach to cybersecurity across industries. Anthropic’s positioning with Claude Mythos for Project Glasswing is two-pronged—for cybersecurity applications including finding vulnerabilities in software, and secondly, for zero-day vulnerabilities that were previously not discovered or known to exist. Also Read: OpenAI, Anthropic, Google unite to combat AI model copying in China The AI company’s Frontier Red Team, a specialised technical group that tests AI models for capabilities and any possible harmful indicators in autonomous or specific behaviour, notes that while the Opus 4.6 model mostly had a near 0% success rate at autonomous exploit development, the Mythos Preview does significantly better. “For example, Opus 4.6 turned the vulnerabilities it had found in Mozilla’s Firefox 147 JavaScript engine—all patched in Firefox 148—into JavaScript shell exploits only two times out of several hundred attempts. We re-ran this experiment as a benchmark for Mythos Preview, which developed working exploits 181 times, and achieved register control on 29 more,” they note, regarding cybersecurity capabilities. Anthropic wants the industry to believe that Claude Mythos is better than anything before it, and that may well be the case. Alongside, selective deployment via a consortium lends credence to the projection of responsibility. But more than anything else, in its entirety, this seems to be an extraordinarily effective narrative starting with the purported leak. ABOUT THE AUTHOR Vishal Mathur Vishal Mathur is Technology Editor for Hindustan Times. When not making sense of technology, he often searches for an elusive analog space in a digital world. Artificial Intelligence Stay updated with the latest Business News on Petrol Price, Gold Rate, Silver Rates, Diesel Prices along with Income Tax Calculator Home / Business / Project Glasswing Aims To Secure The World’s Software, And Anthropic’s Mythology See LessThere’s an irony underlying this situation. An artificial intelligence (AI) model that Anthropic was working on was ‘leaked’ a few days ago. Codenamed Mythos, and capabilities claimed to be “substantially beyond those of any model we have previously trained, this model was discovered after descriptions of the model were found in publicly accessible data following an ‘error’ in the AI company’s content management system. Now, Anthropic says Claude Mythos is so powerful, they’ll not make this model generally available. Instead, it provides the foundation for something called Project Glasswing, to secure every other software.
Where have you heard this before? In 2019, Dario Amodei who is now the CEO and co-founder of Anthropic was still at OpenAI as Vice President of Research. In February 2019, OpenAI had said that its text-generation GPT-2 algorithm was too dangerous for public release. Has the marketing playbook been carried forward with the passage of time? now, this model is the basis for an industry consortium, called Project Glasswing, with an intention to deploy this for cybersecurity
“Instead, we are using it as part of a defensive cybersecurity program with a limited set of partners,” the company details, in the system card for the Claude Mythos Preview released this week. The partners include Apple, Google, Microsoft, Amazon Web Services, Nvidia and Cisco. Anthropic insists the model’s “striking leap in scores on many evaluation benchmarks” is the reason for the advantage over Claude Opus 4.6 across agentic coding, reasoning and computer use benchmarks.
The GPT-2 resemblance seems difficult to ignore. Back then, the rhetoric was more about AI responsibility and restraint. Eventually, GPT-2 was released (small model in Feb 2019, the medium model in May and the larger models in August and November) , relegating the ‘too dangerous to share’ pitch more to the AI marketing folklore, than a pure safety debate. The Project Glasswing revives those memories, as a model that’s claimed to be exceptionally capable and therefore very risky, with a need for controlled usage.
“As part of Project Glasswing, the launch partners listed above will use Mythos Preview as part of their defensive security work; Anthropic will share what we learn so the whole industry can benefit. We have also extended access to a group of over 40 additional organisations that build or maintain critical software infrastructure so they can use the model to scan and secure both first-party and open-source systems,” the company says in a statement.
Anthropic says they’ll be committing up to $100 million worth of usage credits for the Mythos Preview across these efforts—after that, Claude Mythos Preview will be available to participants at $25/$125 per million input/output tokens. Important to note the commercial architecture that’s being created here, with Claude Mythos still very much a product in Anthropic’s scheme of things—scarcity, capabilities, and therefore premium pricing strategies will be at work.
They are bullish about their claims, saying the model found a 27-year-old vulnerability in OpenBSD operating system and a 16-year-old vulnerability in FFmpeg software used to encode and decode video. They also say that the Mythos Preview has already found thousands of high-severity vulnerabilities—including some in every major operating system and web browser.
The partners insist this will be one part of their overall cybersecurity efforts.
“We will take a rigorous, independent approach to determining how to proceed and where we can help,” says Pat Opet, Chief Information Security Officer at JPMorganChase. Heather Adkins, VP of Security Engineering at Google, says they’ll make the Mythos Preview available via the Vertex AI platform and “will continue investing in our leading cybersecurity platform and a culture focused on protecting users.”
Amy Herzog, Amazon Web Services’ Vice President of CSIO, notes that their teams analyse over 400 trillion network flows every day for threats, and AI is crucial for identifying and defending at scale. They’ve applied Claude Mythos Preview to these security operations, and it has helped strengthen the code in critical instances.
The unwritten caveat must be noticed. None of the Project Glasswing partners are fully handing over their cybersecurity apparatus or architecture to the new model. This will be an additional layer, not the entirety. This lends the necessary perspective. Claude Mythos’ claimed scores and capabilities are impressive, but it is unlikely that one AI model will fundamentally restructure the approach to cybersecurity across industries.
Anthropic’s positioning with Claude Mythos for Project Glasswing is two-pronged—for cybersecurity applications including finding vulnerabilities in software, and secondly, for zero-day vulnerabilities that were previously not discovered or known to exist.
Also Read: OpenAI, Anthropic, Google unite to combat AI model copying in China
The AI company’s Frontier Red Team, a specialised technical group that tests AI models for capabilities and any possible harmful indicators in autonomous or specific behaviour, notes that while the Opus 4.6 model mostly had a near 0% success rate at autonomous exploit development, the Mythos Preview does significantly better.
“For example, Opus 4.6 turned the vulnerabilities it had found in Mozilla’s Firefox 147 JavaScript engine—all patched in Firefox 148—into JavaScript shell exploits only two times out of several hundred attempts. We re-ran this experiment as a benchmark for Mythos Preview, which developed working exploits 181 times, and achieved register control on 29 more,” they note, regarding cybersecurity capabilities.
Anthropic wants the industry to believe that Claude Mythos is better than anything before it, and that may well be the case. Alongside, selective deployment via a consortium lends credence to the projection of responsibility. But more than anything else, in its entirety, this seems to be an extraordinarily effective narrative starting with the purported leak.
Source: HindustanTimes
Related Posts: Mumbai-Pune Expressway’s ‘Missing Link’ Project Aims For Guinness World Record Gurugram-Noida Rail Project Aims To Cut Travel Time To 38 Minutes Software used to delete voters in Karnataka’s Aland Airbus software update Adobe to buy software provider Semrush in $1 DGCA Issues Safety Directive For Airbus Models After Software Fix Alert- Read Services surplus cushions trade gap as software Gujarat police to use AI-powered software to manage Rath Yatra crowd US Supreme Court rejects software giant SAP's bid to avoid rival's antitrust suit American Airlines cuts number of jets needing Airbus software fix to 209
As someone who grew up in a border region in the 1980s and early 1990s, where Urdu was the official language of Jammu and Kashmir, and a language close to urban Punjabi served as the native language, television often became a site of small but spirited conflicts among siblings
3 minutes ago
Aquarius Horoscope Today for April 9, 2026: Mars in Aries may bring positive signalsAquarius Horoscope Today: This is a good day for practical progress, planning, clean decisions, and focusing on what can actually be completed. Published on: Apr 09
3 minutes ago
Voting in the Assembly elections in Assam, Kerala and Puducherry began on Thursday. As of 1 pm, the turnout in Assam was 59.6%, Kerala 49.7% and Puducherry 56.8%, according to data from the Election Commission. All constituencies in the two states and the Union Territory are being covered in a
3 minutes ago
‘Oracle’ effect: Mass layoffs force family in US to reverse migrate to IndiaOracle has recently laid off around 30,000 employees globally, including about 12,000 in India. Published on: Apr 09, 2026 2:39 PM IST By Trisha Sengupta Share via Copy link A viral Reddit post has highlighted the
3 minutes ago
Sunetra Pawar set to win Baramati bypoll unopposed as Congress withdraws candidateCongress said the party had initially gone ahead with its campaign after fielding Akash More for the Baramati bypoll. But the party has decided to withdraw now. Updated on: Apr 09
3 minutes ago
Editorial independence is core to our work. Some links may earn us a commission, without influencing our opinions.These straight jeans won’t give up on you even after 10 washes: 5 picks for womenStraight jeans are the quiet overachievers of your wardrobe; they don’t scream for attention
3 minutes ago
BJD president Naveen Patnaik accepted MP Sasmit Patra's resignation as Rajya Sabha parliamentary party leader following the defeat of a supported candidate amid cross-voting. The party suspended six MLAs for "anti-party activities." This follows Patra's earlier resignation from a committee in
3 minutes ago
Amazon has announced that it will stop supporting its Kindle e-reader devices released in 2012 and before, as users scrambled to figure out how their hardware and virtual libraries would be affected by the tech company’s policy. According to Amazon
3 minutes ago
A: The earth formed as a molten world. Around 4 billion years ago, its surface cooled enough for water vapour to condense. Torrential rains then fell for centuries, filling basins to create the first oceans around 3.8 billion years ago. Once liquid water covered large parts of the surface
3 minutes ago
There is something quietly significant about what VinFast is attempting with the VF MPV 7. In a market like India, where electric mobility is still evolving beyond early adopters, the idea of a practical, family-focused electric MPV feels less like a gamble and more like a calculated step forward
3 minutes ago
Think about sitting down for a meal as missiles begin to rain from the sky. Rani, an Indian woman from Gujarat living in Israel, shares soul-stirring footage of life under fire—from abandoned meals to nights spent in cramped bunkers. Rani recently posted an Instagram reel showing what life is
3 minutes ago
Karnataka 2nd PUC Result 2026 Live: The Karnataka School Examination and Assessment Board (KSEAB) declared Class 12 (2nd PUC) results today, April 9, 2026 on karresults.nic.in, kseab.karnataka.gov.in, after delays due to data compilation issues and Election Commission permissions amid the model
3 minutes ago
Israel said it killed a close aide and nephew of Naim Qassem in an overnight strike on Beirut, targeting the leadership circle of the Iran-backed Hezbollah amid escalating regional tensions. Israel said on Thursday it had killed the nephew of Naim Qassem
3 minutes ago
Leading Red Bull engineer Gianpiero Lambiase has agreed to join McLaren from 2028. Briton Lambiase, who is most well known as being Max Verstappen's race engineer, will take on a role supporting McLaren team principal Andrea Stella. The exact details and title of the new position are not yet known
3 minutes ago
In its latest India Equity Strategy Playbook report, the brokerage firm said that the current combination of depressed valuations, improving earnings momentum and cautious investor positioning reflects conditions typically seen near the end of market downturns
3 minutes ago
The Modi government increases natural gas allocation to fertiliser plants amid improving Gulf conditions, ensuring better fertiliser availability and relief for farmers. West Asia war update: In a big development hours after the announcement of ceasefire deal signed between Iran and the United
3 minutes ago
Editorial independence is core to our work. Some links may earn us a commission, without influencing our opinions.Bought a new smart TV? Grab a new Dolby Digital soundbar to complete the home theatre setupDolby Digital soundbars for smart TVs deliver clearer dialogue
3 minutes ago
Changing the shape of your side for a big game will always draw attention. So Arne Slot put himself in the spotlight by setting up Liverpool with a back five against the champions of Europe. The result - a 2-0 defeat by Paris St-Germain in the first leg of their Champions League quarter-final -
3 minutes ago
Hours before US President Donald Trump's expletive-laden threat was to come into effect, both Iran and Washington announced that they had agreed to a two-week ceasefire. In the final hours before this major announcement, Pakistan served as the primary mediator, according to an Axios report
3 minutes ago
Iran's Revolutionary Guard published charts suggesting sea mine placement in the Strait of Hormuz, potentially pressuring the U.S. amid a fragile ceasefire. Oil prices rose on skepticism over the truce, while U.S. forces will remain in the region. Peace talks are expected in Pakistan
3 minutes ago
With mango season already in full swing across India this summer, there’s no better time to bring back the classic aam panna, a refreshing, centuries‑old drink that has cooled homes long before air conditioners arrived. Made from raw or ripe green mangoes, this tart
3 minutes ago
The Bharatiya Janata Party (BJP) may be pitching “parivartan" (change) as the central theme of its campaign across West Bengal, but a parallel call for change is emerging in the hills of Darjeeling led by a new political entrant. The political landscape in the hills is currently dominated by BJP
3 minutes ago
Founder hails employee for being online on wedding day, internet calls it ‘awful’The founder praised the employee’s sense of ownership, adding that he was asked to log off instantly. Published on: Apr 09, 2026 2:48 PM IST By Trisha Sengupta Share via Copy link A founder is facing a wave of
3 minutes ago
IPL RR vs RCB: Rajasthan Royals and Royal Challengers Bengaluru are set to face off in a key IPL match in Guwahati. Rajasthan boast a powerful opening pair, while Bengaluru rely on a strong top order and a revitalised bowling unit. Guwahati: Riding on explosive starts from and Vaibhav Sooryavanshi
3 minutes ago
Shubman Gill fined ₹12 lakh for slow over-rate in GT vs DC IPL 2026 clash. Here’s what happened and why it matters. In a dramatic IPL 2026 clash that went down to the final ball, Gujarat Titans captain Shubman Gill has been fined ₹12 lakh for breaching the IPL Code of Conduct during Match 14
3 minutes ago
KSEAB 12th Result 2026: Karnataka PUC 2 results out at karresults.nic.in, direct link to check hereKSEAB 12th Result 2026 has been declared. The direct link to check Karnataka PUC 2 results is given here. Published on: Apr 09, 2026 2:53 PM IST By HT Education Desk Share via Copy link The Karnataka
3 minutes ago
Saudi Arabia’s Ministry of Hajj and Umrah has introduced a new initiative offering deals and discounts for Hajj and Umrah pilgrims through the Nusuk Card, aiming to improve their overall experience during the pilgrimage. The Nusuk Card serves as an official identification for pilgrims as well as
3 minutes ago
