In News:

A ransomware group named World Leaks allegedly leaked internal documents related to the Kudankulam Nuclear Power Plant (KKNPP) on the dark web, reviving concerns over the cybersecurity of India's critical infrastructure. While the Nuclear Power Corporation of India Limited (NPCIL) clarified that the leaked material was limited to non-critical facilities outside the reactor island, the incident has renewed focus on protecting strategic assets from cyber threats.

About the Recent Data Leak

The leaked documents reportedly included engineering drawings, inspection records, technical reports, meeting minutes and official correspondence. NPCIL stated that these documents did not relate to reactor operations or nuclear safety, as the highly secured reactor island remained unaffected. Nevertheless, the incident highlights that even non-operational networks associated with strategic installations can become targets for ransomware and cyber espionage.

2019 Kudankulam Cyberattack

The present incident recalls the 2019 cyberattack on Kudankulam, one of India's most significant cyber intrusions involving nuclear infrastructure.

The attack was traced to DTrack malware, linked to the North Korea-backed Lazarus Group, a hacking organisation known for cyber espionage and financial attacks. The malware infected the administrative network's domain controller, potentially exposing user credentials and sensitive information. Investigations suggested that attackers were particularly interested in India's thorium-based nuclear programme.

The malware reportedly entered the network through phishing emails containing malicious links sent to serving and retired nuclear scientists using official email accounts. NPCIL later clarified that only the internet-connected administrative network was compromised, while the reactor control systems remained isolated and unaffected.

Air-Gapped Networks

Nuclear facilities rely on air-gapped networks, which are physically isolated from external networks to minimise cyber risks.

They generally operate through two separate networks:

• Operational Technology (OT) Network: Controls reactors, turbines and other critical plant operations.
• Information Technology (IT) Network: Supports administrative functions such as communication, procurement and documentation.

Although air-gapping significantly enhances security, it is not completely foolproof. Malware can still enter through infected USB devices, maintenance equipment, insider threats or phishing attacks. Global incidents such as the Stuxnet attack on Iran's nuclear programme demonstrate that even isolated systems remain vulnerable.

Why Cybersecurity of Nuclear Infrastructure Matters

Nuclear installations constitute Critical Information Infrastructure (CII), where cyberattacks can have consequences far beyond data theft.

Potential risks include:

• Theft of sensitive scientific and engineering information.
• Strategic cyber espionage.
• Disruption of operational systems.
• Erosion of public confidence in nuclear safety.
• Threats to national security, energy security and environmental safety during geopolitical conflicts.

As ransomware groups and state-sponsored cyber actors become increasingly sophisticated, protecting both operational and administrative networks has become an essential component of national security.

Measures to Strengthen Cybersecurity

India has adopted multiple institutional and technical safeguards to improve cyber resilience.

• CERT-In (Indian Computer Emergency Response Team): National nodal agency for responding to cybersecurity incidents.
• National Critical Information Infrastructure Protection Centre (NCIIPC): Established under the Information Technology Act, 2000, to secure critical infrastructure across strategic sectors.
• National Cyber Security Policy, 2013: Provides the national framework for enhancing cybersecurity and cyber resilience.
• Regular cybersecurity audits, penetration testing and continuous network monitoring.
• Deployment of air-gapped operational networks, stronger access controls and improved security protocols for sensitive installations.

Challenges

Despite these measures, several challenges persist. Administrative networks often remain exposed to phishing attacks and insider threats. Growing digitalisation of critical infrastructure increases the attack surface, while sophisticated ransomware groups and state-sponsored hackers continue to evolve their capabilities. Balancing greater private sector participation in strategic sectors with robust cybersecurity standards is another emerging concern.

Way Forward

India needs to adopt a comprehensive Zero Trust Architecture, strengthen cyber hygiene through continuous employee training, enhance AI-enabled threat detection, conduct regular red-team exercises for critical infrastructure, improve coordination between CERT-In, NCIIPC and sectoral agencies, and develop indigenous cybersecurity solutions for strategic installations. Greater investment in cyber resilience, supply chain security and secure operational technologies will be essential to safeguard critical national infrastructure.

In News:

The rapid evolution of Artificial Intelligence (AI) has reached a critical juncture with Anthropic’s unveiling of Claude Mythos. Representing a "step-change" in Large Language Model (LLM) capabilities, Mythos is not merely an incremental update to the Claude family (Haiku, Sonnet, Opus) but a specialized, high-impact system designed for autonomous vulnerability detection. Its emergence highlights the "dual-use" nature of frontier AI, where the same tool capable of fortifying digital defenses can simultaneously serve as a potent weapon for cyber warfare.

Claude Mythos: Technical Leap and Capabilities

Unlike general-purpose models, Claude Mythos is engineered for advanced reasoning and complex problem-solving within software architecture.

1. Vulnerability Detection at Scale: In early testing, Mythos demonstrated the ability to identify thousands of critical security flaws in legacy software, operating systems, and web browsers—some of which had remained undetected by human reviewers for decades.
2. Unprecedented Efficiency: Researchers indicate that Mythos is approximately an order of magnitude faster than existing automated tools. It can analyze vast, complex codebases and generate patches with minimal human supervision.
3. From Coding to Security: While earlier models like Claude Opus showed strong coding proficiency, Mythos transitions from simply writing code to identifying deep-seated structural vulnerabilities in existing infrastructure.

Project Glasswing: The Defensive Strategy

Recognizing the risks of a public release, Anthropic has adopted a strategy of "Restricted Access" through Project Glasswing.

• Consortium-Based Rollout: Instead of a commercial launch, access is limited to a consortium of over 40 major entities, including tech giants like Microsoft, Apple, and Cisco, as well as infrastructure operators.
• The "Defender’s Advantage": The strategic logic is to provide a "head start" to legitimate defenders. By allowing critical infrastructure providers to identify and patch flaws first, the project aims to secure the digital ecosystem before similar capabilities inevitably diffuse to malicious actors.

The Dual-Use Dilemma and Global Risks

The emergence of Mythos underscores a growing concern in the global security landscape:

• Blurring of Offensive and Defensive Lines: A tool that finds a bug to fix it can also be used by a hacker to exploit it.
• Lowering the Entry Barrier: AI-driven tools could allow individuals with limited technical expertise to launch sophisticated "Zero-Day" attacks.
• Speed of Exploitation: By compressing the time between vulnerability discovery and exploit generation, the window for manual human response is shrinking, necessitating AI-on-AI defense mechanisms.

Implications for India’s Cybersecurity Landscape

For India, a global IT hub with a burgeoning digital economy, Claude Mythos presents a unique set of challenges and opportunities:

• Dependence on Global Ecosystems: Since India relies heavily on both foreign platforms and domestically developed software, the vulnerabilities discovered (and patched) via Mythos will have a direct impact on Indian digital resilience.
• The Participation Gap: Currently, no Indian firm is part of the Project Glasswing consortium. This lack of early access may leave Indian-developed software and critical infrastructure exposed to threats that international counterparts have already mitigated.
• Institutional Response:
  1. CERT-In (Indian Computer Emergency Response Team): The national agency is reportedly studying the implications of Mythos to formulate response strategies.
  2. DSCI (Data Security Council of India): Under the aegis of NASSCOM, the DSCI is actively discussing the impact of such models on the domestic IT industry.
  3. Policy Imperative: There is an urgent need for India to enhance its institutional readiness and foster domestic AI-security research to avoid strategic dependence on foreign "defensive" consortiums.

Conclusion

Claude Mythos represents the dawn of autonomous cyber operations. While it offers a revolutionary way to secure the "technical debt" of decades-old legacy code, its existence forces a global rethink of AI safety and democratization. For aspirants and policymakers, the Mythos case study is a reminder that in the digital age, national security is increasingly defined by the ability to master—and regulate—the algorithms that govern the code.

In News:

In 2024, India has witnessed an alarming rise in cybercrime, particularly a new scam called "digital arrests." This type of fraud involves criminals impersonating law enforcement officials to extort money from victims. With more than 92,000 people targeted and ?2,141 crore defrauded from victims, these scams are rapidly becoming a significant concern for the public and law enforcement.

Nature of ‘Digital Arrests’

The modus operandi of digital arrest scams is sophisticated and emotionally manipulative. Cybercriminals contact victims through video calls, often using fake police officers' profiles and official documents to build credibility. They accuse victims of serious crimes such as money laundering or drug trafficking, claiming urgent action is needed to avoid arrest. The scammers create a false atmosphere of fear and urgency, convincing the victim to transfer large sums of money under the pretext of settling legal dues.

A notable example involves Ruchi Garg, who was targeted by scammers posing as police officers, falsely claiming her son was involved in a major scam. She was coerced into transferring ?80,000 before realizing it was a scam. Similar cases have affected hundreds, with perpetrators using AI-generated voices and fake visuals to amplify the deception.

The Growth of Cybercrime in India

Digital arrest scams are part of a broader increase in cybercrime in India. The Indian Cyber Crime Coordination Centre (I4C) has reported a rise in cyber fraud, with financial losses exceeding ?27,900 crore between 2021 and 2024. The most significant sources of these losses include stock trading scams, Ponzi schemes, and digital arrest frauds. As criminals adapt to emerging technologies and use social engineering tactics, the scale and complexity of scams are growing.

The surge in cybercrimes is fueled by vulnerabilities in India's digital landscape. With over 95 crore Internet users, many people, particularly the elderly or less tech-savvy, remain susceptible to such fraud. Cybercriminals often exploit this lack of awareness, combining fear and confusion to manipulate victims.

International Scope and Challenges

One of the challenges in combating digital arrests is the transnational nature of cybercrime. Scams often originate from countries like China, Cambodia, and Myanmar, where "scam compounds" run operations to train individuals in fraudulent techniques. These groups use virtual private networks (VPNs) and encrypted apps to conceal their identities and locations, making it difficult for Indian authorities to trace them.

Moreover, the involvement of mule bank accounts to launder defrauded money complicates investigations. Thousands of such accounts are identified and blocked regularly, but the flow of money continues through multiple channels, including cryptocurrencies.

Government Efforts and Preventive Measures

To address the growing menace of digital frauds, the Indian government has initiated several measures. The I4C, launched in 2020, aims to strengthen the response to cybercrimes by coordinating with various law enforcement agencies. The National Cyber Crime Reporting Portal allows citizens to report cyber fraud, while real-time alerts are sent to banks to prevent financial losses.

Additionally, the Cyber Crime Coordination Centre and other initiatives like Cyber Surakshit Bharat and CERT-In are working to enhance cybersecurity awareness and support victims. The Digital Personal Data Protection Act, 2023, also aims to regulate data security, which can reduce the sale of personal data on the dark web, a key enabler of these scams.

Conclusion

‘Digital arrests’ exemplify the evolving nature of cybercrimes in India. As digital threats become more complex and widespread, it is essential for citizens to remain vigilant and informed. Effective law enforcement, technological innovations, and public awareness are critical to reducing the impact of these scams and safeguarding the digital economy.

Why is it in the News?

As generative AI technology becomes more prevalent, safeguarding consumers' ability to navigate digital environments securely has become increasingly imperative.

Context:

• In recent times, the integration of generative artificial intelligence (AI) across industries has significantly transformed operational processes.
• However, this rapid advancement has also led to the emergence of new cyber threats and safety concerns.
• With incidents such as hackers exploiting generative AI for malicious purposes, including impersonating kidnappers, it is evident that a comprehensive analysis and proactive approach are required to address and mitigate the potential risks associated with this technology.
• A study by Deep Instinct revealed that 75% of professionals observed a surge in cyberattacks over the past year, with 85% attributing this escalation to generative AI.
• Among surveyed organizations, 37% identified undetectable phishing attacks as a major challenge, while 33% reported an increase in the volume of cyberattacks.
• Additionally, 39% of organizations expressed growing concerns over privacy issues stemming from the widespread use of generative AI.

Significant Impact of Generative AI & Growing Cybersecurity Challenges:

• Transformative Impact: Generative AI has revolutionized various sectors like education, banking, healthcare, and manufacturing, reshaping our approach to operations.
  • However, this integration has also redefined the landscape of cyber risks and safety concerns.
• Economic Implications: The generative AI industry's projected contribution to the global GDP, estimated between $7 to $10 trillion, underscores its significant economic potential.
  • Yet, the development of generative AI solutions, such as ChatGPT introduced in November 2022, has introduced a cycle of benefits and drawbacks.
• Rising Phishing and Credential Theft: An alarming surge of 1,265% in phishing incidents/emails and a 967% increase in credential phishing since late 2022 indicates a concerning trend.
  • Cybercriminals exploit generative AI to craft convincing emails, messages, and websites, mimicking trusted sources to deceive unsuspecting individuals into divulging sensitive information or clicking on malicious links.
• Emergence of Novel Cyber Threats: The proliferation of generative AI has expanded the cyber threat landscape, enabling sophisticated attacks.
  • Malicious actors leverage AI-powered tools to automate various stages of cyber-attacks, accelerating their pace and amplifying their impact.
  • This automation poses challenges for detection and mitigation, making attacks more challenging to thwart.
• Challenges for Organizations: Organizations across sectors face escalating cyber threats, including ransomware attacks, data breaches, and supply chain compromises.
  • The interconnected nature of digital ecosystems exacerbates the risk, as vulnerabilities in one system can propagate to others, leading to widespread disruption and financial losses.
  • Additionally, cybercriminals' global reach and anonymity pose challenges for law enforcement and regulatory agencies.

The Bletchley Declaration: Addressing AI Challenges

• Global Significance: The Bletchley Declaration represents a pivotal global initiative aimed at tackling the ethical and security dilemmas associated with artificial intelligence, particularly generative AI.
  • Named after Bletchley Park, renowned for its British code-breaking endeavours during World War II, the declaration embodies a collective resolve among world leaders to shield consumers and society from potential AI-related harms.
• Acknowledgement of AI Risks: The signing of the Bletchley Declaration at the AI Safety Summit underscores the mounting awareness among global leaders regarding AI's inherent risks, notably in the cybersecurity and privacy realms.
  • By endorsing coordinated efforts, participating nations affirm their dedication to prioritizing AI safety and security on the international agenda.
• Inclusive Engagement: The Bletchley Declaration's inclusive nature is evident in the involvement of diverse stakeholders, including major world powers like China, the European Union, India, and the United States.
  • By fostering collaboration among governments, international bodies, academia, and industry, the declaration facilitates cross-border and cross-sectoral knowledge exchange, essential for effectively addressing AI challenges and ensuring equitable regulatory frameworks.
• Consumer Protection Focus: At its heart, the Bletchley Declaration underscores the imperative of safeguarding consumers against AI-related risks.
  • Participating countries commit to formulating policies and regulations that mitigate these risks, emphasizing transparency, accountability, and oversight in AI development and deployment.
  • Additionally, mechanisms for redress in cases of harm or abuse are prioritized.
• Ethical AI Promotion: A core tenet of the Bletchley Declaration is the promotion of ethical AI practices.
  • Participating nations pledge to uphold principles of fairness, accountability, and transparency in AI development and usage, striving to prevent discriminatory or harmful outcomes.
  • This commitment aligns with broader endeavours to ensure responsible AI deployment for the betterment of society.

Alternative Measures for AI Risk Mitigation:

• Institutional-Level Strategies: Governments and regulatory bodies can enact robust ethical and legislative frameworks to oversee the development, deployment, and utilization of generative AI technologies.
  • These frameworks should prioritize consumer safeguarding, transparency, and accountability, all while fostering innovation and economic prosperity.
  • Furthermore, the integration of watermarking technology can aid in the identification of AI-generated content, empowering users to discern between authentic and manipulated information.
  • This proactive approach can substantially mitigate the prevalence of misinformation and cyber threats stemming from AI-generated content.
• Continuous Innovation and Adaptation: Sustained investment in research and development is imperative to proactively address emerging cyber threats and devise innovative solutions to counter them.
  • By bolstering support for cutting-edge research in AI security, cryptography, and cybersecurity, governments, academia, and industry can drive technological progress that fortifies cybersecurity resilience and mitigates the inherent risks associated with generative AI.

Conclusion

Effectively tackling the challenges presented by generative AI demands a comprehensive strategy encompassing regulatory, collaborative, and educational efforts across institutional, corporate, and grassroots domains. Through the enactment of robust regulatory frameworks, stakeholders can collaboratively mitigate the risks posed by AI-driven cyber threats, fostering a safer and more secure digital environment for all.