Anthropic’s Glasswing Sharpens Risks for Indian IT, Global Banks

News

1. Anthropic’s Glasswing Sharpens Risks for Indian IT, Global Banks

2. Cisco Eyes Astrix Deal as AI Agents Trigger New Cybersecurity Risks

3. Anthropic Begins Project Glasswing to Hunt Software Flaws With AI

4. AP Positions Amaravati as Quantum Hub with Launch of Testing Facility

5. Cisco Targets AI Reliability Gap in Planned Galileo Deal

6. Intel, Google Expand AI Infra Partnership with Focus on CPUs

Anthropic’s Project Glasswing is emerging as a potential inflection point for India’s $250 billion IT services industry and the global financial system, with analysts warning that advances in artificial intelligence could reshape software economics and cybersecurity risk.

Anthropic’s unreleased Mythos model represents a “step-jump in benchmark performance across software engineering tasks,” Kotak Institutional Equities said in a report, raising “near- to medium-term disruption risks for IT services,” particularly for companies reliant on application development.

Application development, or the design, coding, and maintenance of software systems, has traditionally required large teams of engineers and accounts for a significant share of revenue at Indian IT firms.

As AI systems become capable of performing many of these tasks, clients may demand faster delivery and lower costs, potentially compressing margins and slowing revenue growth.

Kotak estimated that generative AI could already create a 3%-3.5% annual growth headwind for the sector. With Mythos-level capabilities, the brokerage said that estimate “may no longer be conservative.”

“The realization of similar improvements in real-world scenarios risks turning our estimate from prudent to practical,” the report said.

From Coding to Cybersecurity

Project Glasswing is built around Anthropic’s Claude Mythos Preview, a frontier AI model designed to identify and help fix software vulnerabilities, or weaknesses in code that attackers can exploit to gain unauthorized access to systems.

Anthropic has restricted access to Mythos Preview to launch partners using it for defensive security work and to more than 40 additional organizations that build or maintain critical software infrastructure, according to the company.

The initiative also points to a shift in cybersecurity economics. Traditionally, companies have relied on penetration testing, a process in which security experts simulate cyberattacks to uncover vulnerabilities.

Advances in AI systems like Mythos could automate more of this discovery process, potentially reducing the need for large teams of human testers and putting pressure on pricing for such services.

Another key concept is the zero-day vulnerability, a flaw in software that can be exploited before developers have time to issue a fix.

Mythos’s reported ability to identify such vulnerabilities at scale has raised concerns that the gap between discovery and exploitation could shrink, increasing systemic cyber risk.

Regulatory Concerns

The implications of Glasswing extend beyond enterprise productivity.

Reports of discussions among policymakers and banking leaders in the US and UK point to growing concern that highly capable AI systems could expose vulnerabilities in critical financial infrastructure. 

Reuters reported that UK authorities, including the Bank of England, the Financial Conduct Authority, and the Treasury, were coordinating with the National Cyber Security Centre to assess the risks and prepare briefings for major financial institutions. 

Such risks are particularly significant for banks, which rely heavily on complex and interconnected software systems.

For financial institutions, the harder task is remediation, meaning validating the risk, developing patches, and deploying fixes without disrupting operations.

As AI accelerates vulnerability discovery, remediation capacity may become a bigger constraint in cybersecurity.

Kotak said the effects of AI-driven automation would be uneven across service lines. While efficiencies are expected across the industry, application development is likely to face the sharpest disruption because of its reliance on repetitive coding tasks that can be automated.

Kotak also pointed to areas that could offset some of the pressure. As companies step up spending on AI, demand is likely to rise for work such as modernizing legacy systems, rebuilding data infrastructure so enterprise information can be used effectively by AI tools, and helping clients secure and fix systems as machine-led vulnerability discovery becomes more common.

The brokerage also sees scope for new revenue from broader generative AI deployments across business operations. Those opportunities play to areas where Indian IT firms have traditionally been stronger, including large-scale systems integration, domain-led consulting, and the management of complex technology transitions.

Project Glasswing could begin to redefine where value is created in both technology services and cybersecurity.

As AI begins to commoditize routine coding and vulnerability discovery, more value could shift toward higher-order work such as governance, risk management, and large-scale remediation.

The initiative may signal a repricing of the IT services model. Companies that rely heavily on labor-intensive execution may face margin pressure, while those that pivot toward AI-led transformation and security services could find new growth avenues.

Rather than signaling an immediate downturn for Indian IT, Project Glasswing may mark the beginning of a structural transition. Kotak’s analysis points to a dual reality: near-term disruption in traditional service lines alongside longer-term opportunities driven by enterprise AI adoption.