Information & Communication Technology

What is Distributed Denial of Service (DDoS) attack?

Context: A highly anticipated livestream conversation between former US President Donald Trump and X owner Elon Musk was severely disrupted by a massive Distributed Denial of Service (DDoS) attack recently. The incident forced X to scale down the live audience and delay the interview.

Distributed Denial of Service (DDoS) Attack

Distributed Denial of Service (DDoS) Attack
  • Denial of Service attack is a type of cyber-attack meant to shut down a machine or network, making it inaccessible to its intended users. A DoS attack originates from a single source, typically one computer or network connection.
  • DDoS (Distributed Denial of Service) is a type of cyberattack where multiple compromised computers, often part of a botnet, are used to flood a targeted server, website, or network with an overwhelming amount of traffic. This flood of traffic overwhelms the target's resources, making it difficult or impossible for legitimate users to access the service.
AspectDenial of Service (DoS)Distributed Denial of Service (DDoS) 
Source of attack Single source (one computer or network)Multiple sources (often thousands of compromised devices) 
Scale of attack Smaller, limited by the capabilities of a single machine Large-scale, leveraging a botnet to amplify the attack 
Complexity Relatively simple to execute and mitigate More complex to execute and significantly harder to mitigate 
Detection Easier to detect due to traffic from a single source Harder to detect due to distributed nature, making it difficult to distinguish between legitimate and malicious traffic  
Impact Can causes disruption but typically less severe Can cause significant disruption, often resulting in widespread outages 
Mitigation Easier to mitigate by blocking the attacking IP address or source Difficult to mitigate due to traffic coming from numerous sources; requires sophisticated defence mechanisms. 

How DDoS Works?

  • Botnet Creation: Attackers compromise and control a large number of devices (computers, IoT devices, etc.) by exploiting vulnerabilities or spreading malware. These devices become part of a botnet—a network of infected devices controlled by the attacker.
  • Attack Initiation: The attacker commands the botnet to send a massive number of requests or data packets to the target server or network simultaneously.
  • Overloading the Target: The sheer volume of traffic overwhelms the target's infrastructure, consuming its bandwidth, processing power, or memory. This can cause the server to slow down significantly or crash entirely, denying access to legitimate users.
  • Disruption of Services: As a result of the overload, the targeted service becomes unavailable or unresponsive, leading to downtime, loss of revenue, and damage to the organisation’s reputation.
types of Distributed Denial of Service (DDoS) attack
Image source: geeksforgeeks

Impact of DDoS Attacks

  • Service Outages: Legitimate users cannot access the targeted service.
  • Financial Losses: Downtime can lead to lost revenue, especially for e-commerce platforms and online services.
  • Reputation Damage: Repeated attacks can erode trust in the organisation's ability to secure its services.
  • Mitigation Costs: Organisations may need to invest in DDoS protection solutions, which can be expensive.

Mitigation Techniques

  • Traffic Filtering: Using firewalls and intrusion detection systems to filter out malicious traffic.
  • Rate Limiting: Limiting the number of requests a server will accept from a particular IP address within a certain timeframe.
  • Content Delivery Networks (CDNs): Distributing traffic across multiple servers in different locations to reduce the impact of the attack.
  • DDoS Protection Services: Services like Cloudflare, AWS Shield offer protection against DDoS attacks by absorbing and filtering malicious traffic.

What is the Internet Archive?

Context: Internet Archive (IA) is embroiled in a major legal challenge as it faces off against traditional publishers accusing it of copyright violations. The free digital library is fighting the forced removal of around half a million books from its platform, which it argues functions like a library. 

Internet Archive

Internet Archive
  • It is an American nonprofit digital library founded in 1996 by Brewster Kahle.
  • It provides free access to collections of digitized materials including websites, software applications, music, audiovisual, and print materials.
  • The Archive also advocates for a free and open Internet.  
  • Its mission is to provide ‘universal access to all knowledge’. 
  • The Internet Archive allows the public to upload and download digital material to its data cluster, but the bulk of its data is collected automatically by its web crawlers, which work to preserve as much of the public web as possible.
  • Its web archive, the Wayback Machine, contains hundreds of billions of web captures.
  • The Archive provides specialized services relating to the information access needs of the print-disabled. Publicly accessible books were made available in a protected Digital Accessible Information System (DAISY) format. 
  • DAISY is designed as a complete audio substitute for print material. 

Case against Internet Archive: 

  • Many traditional publishers have alleged that Internet Archive violated their copyrights and illegally made their books available to the public, by scanning physical copies and distributing the digital files.
  • Traditional publishers were against IA’s temporary ‘National Emergency Library’ (NEL) initiative that it launched during the COVID-19 pandemic. This was to allow more users to access the e-books in its collection while physical libraries were locked down.
    • During the NEL, IA lifted the technical controls enforcing its one-to-one owned-to-loaned ratio and allowed up to ten thousand patrons at a time to borrow each e-book on the Website.
    • In general, IA uses a system known as ‘controlled digital lending to limit the number of people who can access an ebook.
    • It ended its emergency library system after being hit with the lawsuit.
    • Internet Archive used the doctrine of fair use to defend itself in the case, but this did not hold up. 
  • Hachette vs Internet Archive Case (2020): 
    • Traditional publishers Hachette, HarperCollins, Wiley, and Penguin Random House sued Internet Archive.
    • In 2023, an order was issued in favour of the publishers.
    • The order stated: IA’s Website includes millions of public domain e-books that users can download for free and read without restrictions. However, the Website also includes 3.6 million books protected by valid copyrights. 

Why are books being removed?

  • As a result of the lawsuit, IA was forced to remove over half a million books from its database.
  • While IA identifies itself as a library, it has been compared to a shadow library or a piracy database by traditional publishers, who disagree with its ‘controlled digital lending’ approach.
  • Despite the removal, however, the Internet Archive is still home to a rich collection.
    • It still contains 835 billion web pages, 44 million books and texts, 15 million audio recordings, 10.6 million videos, 4.8 million images, and 1 million software programs. 
    • Live concerts and television programs also make up part of this collection.

Wayback Machine: 

  • The Wayback Machine is a digital archive of the World Wide Web founded by the Internet Archive in 1996 and launched to the public in 2001, it allows the user to go ‘back in time’ to see how websites looked in the past.
  • The Wayback Machine was created as a joint effort between Alexa Internet (owned by Amazon.com) and the Internet Archive. 
  • Hundreds of billions of web sites and their associated data (images, source code, documents, etc.) are saved in a database.
  • There is a good chance of finding content such as old websites that no longer exist today, earlier versions of existing websites, deleted social media posts, archived versions of paywalled articles, and archived versions of content that is blocked or censored in some jurisdictions.
  • Wayback Machine is useful for personal research or to access information sources, but users should be cautious about relying on the data obtained through such sources, as the saved information can sometimes be outdated or inaccurate.
  • This has created more than 28 years of web history accessible through the Wayback Machine. 
  • The platform claims users can explore over 866 billion saved web pages through its own search service.
  • ‘Archive-It’ program identifies important web pages on the Internet Archive’s website.
    • Archive-It: Created in early 2006, Archive-It is a web archiving subscription service that allows institutions and individuals to build and preserve collections of digital content and create digital archives.
    • Archive-It allows the user to customize their capture or exclusion of web content they want to preserve for cultural heritage reasons.
    • Through a web application, Archive-It partners can search, catalogue, manage, browse, and view their archived collections. 
    • Periodically, the data captured through Archive-It is indexed into the Internet Archive's general archive.
  • Not all web sites are available because many web site owners choose to exclude their sites. 

Using firewall to block Internet access

Context: Pakistani media outlets reported that the country is planning to implement a Chinese-style firewall to block users from accessing social media platforms. Even those using Virtual Private Networks (VPN) to access the Internet are likely to be hit by this firewall. 

Digital firewall

image 34
  • A digital firewall is a network security system that monitors and controls incoming and outgoing network traffic based on a set of predefined security rules. 
    • The security tools can stop online traffic from reaching certain sites. They can prevent malicious actors from targeting individual users’ computer systems or home networks, and even thwart cyber threats on specific websites.
    • Large firewalls (like the Great Firewall of China) can be used to prevent people from accessing large sections of the Internet i.e, websites, social media sites or information gathering platforms. 
  • Firewalls can be implemented as hardware devices, software running on a computer, or as a cloud-based service. 

How does a digital firewall work? 

  • Packet Inspection: The firewall examines each data packet (a unit of data transmitted over a network) that passes through it. It looks at the header information of the packet, which includes the source and destination IP addresses, port numbers, and protocol type (e.g., TCP, UDP).
  • Access Control Lists (ACLs): The firewall compares the packet information against a set of predefined rules known as an Access Control List. These rules specify which traffic is allowed to pass through the firewall and which traffic should be blocked.
  • Filtering Decisions: Based on the ACL rules, the firewall makes a decision on whether to allow the packet to pass through or to block it. 

Why is it used?

  • Firewall presents a viable alternative to authoritarian nation states who wish to replace a larger free Internet with a controlled intranet.
    • When used by nation-states, a firewall restricts activists, journalists, dissidents, and regime critics from obtaining information critical of the government. 
    • Internet shutdowns and social media blocks also prevent governments or military authorities from being held accountable during periods of civil unrest and violence.
  • However, setting up and maintaining a firewall is expensive. The firewalls require constant monitoring to thwart bad actors and fix security vulnerabilities.

Right to Internet in India: 

  • In the Anuradha Bhasin v/s Union of India 2020, the Supreme Court ruled that the right to freedom of speech and expression and the freedom to practise any profession or carry on any occupation, trade or business over the internet are protected under Article 19 (1) (a) and Article 19 (1) (g) of the Indian Constitution, respectively. 
  • This essentially means that the court recognised internet access as a fundamental right, integral to a democratic society for its proper functioning.

Phi-3-mini: AI based small language models (SLMs)

Context: Microsoft unveiled the latest version of its lightweight Al model, the Phi-3-mini, reportedly the first among three small language models (SLMs) that the company plans to release. 

More about news: 

  • Phi-3-mini has 3.8 billion parameters (a measure of the size and complexity of an Al model) and is trained on a data set that is smaller in comparison to LLMs such as Open Al's GPT-4.
  • The amount of conversation that an AI can read and write at any given time is called the context window and is measured in tokens.
  • It is the first model in its class to support a context window which helps an Al model recall information during a session of up to 128,000 tokens, with little impact on quality. A token is the fundamental unit of data used by a language model to process and generate text.
  • Phi-3-mini requires less computational power and has much better latency.

About Small language models (SLMs):

  • SLMs are compact versions of large language models (LLMS), which can comprehend and generate human language text. 
  • The new model expands the selection of high-quality models for customers, offering more practical choices as they build generative Al applications.
  • It is more streamlined versions of large language models. When compared to LLMs, smaller AI models are also cost-effective to develop and operate, and they perform better on smaller devices like laptops and smartphones.
  • LLMs are trained on massive general data, while SLMs stand out with their specialisation. Through fine-tuning, SLMs can be customised for specific tasks - achieving accuracy and efficiency in the process. Most SLMs undergo targeted training, which demands considerably less computing power and energy compared to LLMs.
  • SLMs also differ from LLMs with reference to inference latency, which is the time taken for a model to make predictions or decisions after receiving input.
  • Their compact size allows for quicker processing, making them more responsive and apt for real-time applications such as virtual assistants and chat-bots and their cost makes them appealing to smaller organisations and research groups.
  • SLMs are highly versatile, useful in applications ranging from sentiment analysis to code generation. Their compact size and efficient computation also make them ideal for use on edge devices and in resource-limited settings.
  • Most popular SLMs: Llama 2 developed by Meta AI, Mistral and Mixtral, Microsoft’s Phi and Orca, Alpaca 7B and StableLM.

Read more about Large Language Models (LLMs):

What is Doxxing?

Context: Issues of Doxxing have become common leading to violation of privacy of citizens.

Doxxing and cyber crime

Doxxing and cyber crime
  • Doxxing is the act of publicizing someone's private personal information (Personally Identifiable Information or PII) online without their consent, often with malicious intent. PII can include home address, phone number, employment details, financial information, etc.
  • Doxxing is a serious cyber-crime that can endanger the victim's physical safety, emotional well-being, employment, and reputation. It disproportionately affects vulnerable groups like women, children and LGBTQ+ individuals.
  • In 2023, India emerged as the primary target for cybercriminals, accounting for 13.7% of all attacks. The most common form of cyber-crime in India is financial fraud, which accounted for 75% of cyber-crime in India between 2020 and 2023.

Laws and Regulations

  • In India, doxxing victims can file complaints through the National Cyber Crime Reporting Portal. They have a legal right to file an FIR.
  • Social media companies operating in India are bound by the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. These rules require platforms to appoint grievance officers, resolve complaints within fixed time periods, and adhere to a code of ethics.
  • Relevant laws include the Information Technology Act, 2000 which penalizes cyber crimes, and the Indian Penal Code which has provisions related to criminal intimidation, stalking, defamation, etc.

Role of Technology Companies

  • Major platforms like Facebook, Instagram, YouTube, Twitter (X) and Reddit have policies prohibiting doxxing and mechanisms to report such content for removal.
  • However, enforcement is often inconsistent and platforms have faced criticism for not doing enough to protect users, especially marginalized communities.
  • Meta's Oversight Board has recommended stricter doxxing rules, noting the disproportionate impact on vulnerable groups and potential for serious offline harms.
  • Discord has updated its community guidelines to treat doxxing and harassment as distinct harms, reflecting an evolving understanding of these issues among platforms.

Safety Measures for Users

  • Use strong, unique passwords and enable two-factor authentication.
  • Be cautious about posting photos/videos that may reveal identifiable locations or sensitive information.
  • Consider potential risks before posting controversial content.
  • If doxxed, document evidence in an incident log, report to platforms and authorities, and lean on support systems.
  • Secure Personally Identifiable Information (PII) and take steps to remove any compromised data from the internet.
  • More than 90% of web users are concerned about doxxing today, and 73% have limited what they share online to avoid being doxxed.

Psychological and Emotional Impact

  • Doxxing can be an extremely traumatic experience, inducing fear, anxiety, depression and even PTSD in victims.
  • The importance of self-care and seeking support from friends, family, and mental health professionals in coping with the aftermath of doxxing.
  • Building resilience and not succumbing to shame or weakness is crucial, as the fault lies entirely with the perpetrator and not the victim.

Government Initiatives

  • In 2024, the Indian government is working on a new set of rules to criminalize deliberate misinformation and doxing as part of legislation to govern the country's digital ecosystem. This legislation, known as the Digital India Act, is set to replace the IT Act, of 2000, which is over two decades old.
  • The Indian Computer Emergency Response Team (CERT-In) is the national agency that deals with cyber security threats.
  • However, experts have pointed out the need for a comprehensive data protection law, reforms in the criminal justice system, and more public awareness initiatives.

International Perspective

  • Doxxing is a global issue, and many countries have taken steps to address it through legislation and policy measures.
  • In the United States, doxxing can be prosecuted under various federal and state laws, including the Computer Fraud and Abuse Act, the Interstate Communications Statute, and the Cyberstalking Statute.
  • The European Union's General Data Protection Regulation (GDPR) provides a framework for protecting personal data and can be used to address doxxing incidents.
  • Comparing India's approach to doxxing with that of other countries can provide valuable insights and ideas for strengthening the legal and policy framework.

Challenges in Enforcement

  • Despite the existence of laws and regulations, enforcing them in doxxing cases can be challenging due to the anonymous and borderless nature of the internet.
  • Investigating doxxing incidents often requires technical expertise and cooperation between multiple stakeholders, including law enforcement agencies, social media platforms, and internet service providers.
  • The transnational nature of many doxxing cases can create jurisdictional issues and complicate the process of gathering evidence and prosecuting offenders.
  • Addressing these challenges requires capacity building, international cooperation, and the development of specialized units within law enforcement agencies.

Role of Civil Society and Media

  • Civil society organizations and media outlets can play a crucial role in raising awareness about doxxing, providing support to victims, and advocating for stronger legal and policy measures.
  • NGOs and helplines can offer counseling, legal assistance, and other services to individuals who have been doxxed, helping them cope with the psychological and practical consequences.
  • Media coverage of doxxing incidents can help bring attention to the issue, shape public opinion, and put pressure on policymakers and platforms to take action.
  • However, media outlets also have a responsibility to report on doxxing cases in an ethical and sensitive manner, avoiding sensationalism and respecting the privacy and dignity of victims.

Future Trends and Emerging Issues

  • As technology evolves, new forms of doxxing and online harassment may emerge, posing additional challenges for individuals, platforms, and policymakers.
  • The increasing use of artificial intelligence and machine learning in content moderation and user profiling may have implications for doxxing prevention and response.
  • The growing importance of data protection and privacy rights may lead to the development of new legal and regulatory frameworks that could impact the handling of doxxing cases.
  • Anticipating and responding to these future trends will require ongoing research, innovation, and collaboration among stakeholders.

US-UK AI Safety Testing Agreement

Context: The US-UK agreement represents a major step towards international cooperation in ensuring the safe and responsible development of AI technologies. It highlights the shared commitment of both nations to address the challenges and risks associated with advanced AI systems.

The agreement reflects the increasing global awareness of the potential risks and benefits of AI and the necessity for collaborative efforts to guide its development. It underscores the importance of establishing international standards and guidelines for AI safety, security, and ethics.

AI Safety and Security

  • US-UK Agreement
    • The agreement facilitates the sharing of critical information on AI capabilities, risks, and best practices between the US and the UK.
    • It promotes the alignment of approaches to ensure the safe deployment of AI systems and enables joint testing exercises to assess the performance and reliability of AI models.
    • The agreement also encourages personnel exchanges between AI Safety Institutes to foster collaboration and knowledge sharing.
    • Follows commitments made at the Bletchley Park AI Safety Summit (2023)
  • Importance of AI Safety
    • AI safety measures are crucial to address the potential risks posed by advanced AI systems, such as algorithmic bias, privacy violations, and security vulnerabilities.
    • Ensuring the safety of AI systems is essential to prevent unintended consequences and protect individual rights and societal values.
    • Responsible AI development involves transparency, accountability, and the incorporation of human oversight and control.
  • Potential Impact of the Agreement
    • The US-UK agreement sets the stage for enhanced global cooperation in AI safety and security, encouraging other nations to follow suit and collaborate on this critical issue.
    • The agreement serves as a model for other nations to emulate, inspiring them to forge similar partnerships and prioritize AI safety in their own AI development efforts.
    • The agreement recognizes the potential risks of AI in spreading misinformation and undermining election integrity, and seeks to develop strategies to counter these threats.

AI Regulation and Policy

  • US Efforts
    • The National Telecommunications and Information Administration (NTIA) in the US has initiated a consultation process to gather insights on the risks, benefits, and potential policy implications of open-source AI models and dual-use foundation models.
    • President Biden issued an executive order in 2023 that outlines the US government's commitment to ensuring the safe and responsible deployment of AI systems.
    • In 2022, the White House released a Blueprint for an AI Bill of Rights, which sets forth principles and guidelines for protecting individual rights and promoting the responsible use of AI.
  • EU AI Act
    • The proposed European Union AI Act seeks to establish comprehensive safeguards on the use of AI systems, with specific provisions for high-risk applications such as law enforcement.
    • The act aims to ensure that AI systems are transparent, explainable, and subject to human oversight, while also empowering consumers to challenge decisions made by AI systems.
    • The EU AI Act recognizes the potential for AI misuse and seeks to establish clear accountability mechanisms for AI developers and deployers.
  • India's Approach
    • India's Ministry of Electronics and Information Technology (MeitY) has issued evolving advisories on the deployment of AI systems in the country.
    • The advisory, issued on March, 2024, directed intermediaries to label any under-trial/unreliable artificial intelligence (AI) models, and to secure explicit prior approval from the government before deploying such models in India.
    • The Indian Government is developing an AI regulation framework, set for release in mid-2024, with the intention of harnessing AI for economic growth and addressing potential risks and harms.

Open-Source AI Models and Implications

  • Prominent Examples
    • Meta has released Code Llama 70B, the largest and best-performing model in the Code Llama family. Code Llama is a state-of-the-art large language model (LLM) capable of generating code, and natural language about code, from both code and natural language prompts.
    • OpenAI's ChatGPT has been released through a controlled API and product-based approach.
    • Dual-Use Foundation Models with widely available weights, enabling both beneficial and malicious applications
  • Implications for Innovation and Competition
    • In 2024, open-source pretrained AI models have gained significant traction, empowering businesses to accelerate growth by combining these models with private or real-time data.
    • Generative AI challenges a core tenet of traditional intellectual property frameworks: only works created by humans are protected by copyright laws.
    • Emerging use cases around generative AI are disrupting traditional views of creativity, authorship, and ownership and pushing the boundaries of copyright law.
    • In 2024, open-source technology faces increased scrutiny as its prolific use, including in proprietary coding, raises the need for pervasive security screening.

Implications for India

  • The AI advisory in India emphasizes transparency, content moderation, consent mechanisms, and deepfake identification to ensure responsible AI deployment and safeguard electoral integrity.
  • AI presents significant opportunities for economic growth in India. The AI industry is estimated to grow year-over-year at a CAGR of 30%. India's AI market is growing at a CAGR of 25-35% and is projected to reach around $17 billion by 2027.
  • However, the adoption of AI technologies may lead to job displacement in certain sectors. As per market trends, more than 16 million working employees in India will need reskilling and upskilling due to AI's influence by 2027.
  • AI technologies have the potential to enhance law enforcement capabilities in India. However, the use of AI in law enforcement also poses risks, such as bias, privacy violations, and potential misuse of power.
  • An updated toolkit for responsible AI practices in law enforcement has been released by INTERPOL and UNICRI in 2024.
  • The NITI Aayog released an approach paper that explores the various ethical considerations of deploying AI solutions in India.

Opportunities for India

  • The digital divide in India is being addressed. There were 751.5 million internet users in India at the start of 2024, when internet penetration stood at 52.4 percent. Initiatives like BharatNet aim to bridge the digital divide and potentially lead to a major positive shift.
  • The Cabinet has approved the comprehensive national-level IndiaAI mission with a budget outlay of Rs.10,371.92 crore. The IndiaAI mission will establish a comprehensive ecosystem catalyzing AI innovation through strategic programs and partnerships across the public and private sectors.
  • India holds a prominent global position in AI skill penetration and talent concentration, showcasing a strong base of AI professionals. There were 4.16 lakh AI professionals, poised to meet the increasing demand expected to reach 1 million by 2026.
  • AI-driven platforms deliver insights to farmers on topics like disease risks, yield forecasts, labor needs, crop protection, weather impacts on crops, and harvest windows.
  • AI has been used thoughtfully by educators to support learning and to give them "time back" in their day. AI applications in education will be overwhelmingly administrative.

Way Forward for India

  • Developing a National AI Strategy
  • Establishing a dedicated AI governance framework and regulatory body
  • Allocating resources and creating incentives for AI research and innovation
  • The Prime Minister, Shri Narendra Modi inaugurated the Global Partnership on Artificial Intelligence (GPAI) Summit. GPAI is a multi-stakeholder initiative with 29 member countries aiming to bridge the gap between theory and practice on AI.
  • India is the lead chair of GPAI in 2024.

Countries like Japan, Rwanda, Benin, Egypt, Morocco, Mauritius, Tunisia, Sierra Leone, and Senegal have developed comprehensive AI strategies and governance frameworks. The Hiroshima AI Process was launched by the G7 under Japan's presidency in May 2023, with the aim of promoting safe, secure, and trustworthy AI.

BhashaNet Portal launched for Promoting Universal Acceptance of Internet

Context: Ministry of Electronics and Information Technology (Meity) and National Internet Exchange of India (NIXI) have launched the BhashaNet Portal on the Universal Acceptance Day.

About BhashaNet Portal

BhashaNet Portal has been developed by NIXI and MEITY to develop an ecosystem to enable citizen to easily create, communicate, transact, process and retrieve information with ease in digital medium without language barrier.

  • Encouraging use of local language website name and email id.
  • Promoting awareness of local language url and email id.
  • Developing policies and regulations.
  • Supporting technical collaboration.
  • Engagement of website owners, web-developer community, web security experts.
image 73

Internationalized Domain Names (IDNs)

  • These enable people to across the world to use domain names in local languages and scripts. IDNs are possible in all Indian languages since they are formed using characters from different scripts such as Hindi, Bengali, Gujrati or Tamil.
  • Under the BhasaNet Portal, .Bharat is available in 22 Scheduled Languages.

Email address internationalization (EAI)

  • It is the process of allowing email addresses to use non-ASCII characters, such as those used in languages like Hindi, Marathi, Bengali, Gujarati or Tamil etc. in addition to traditional ASCII characters used in English-based email addresses.
  • EAI uses Unicode encoding standard to represent non-ASCII characters in email addresses.
  • Requires changes to the way that email is handled by both email clients and email servers to email clients to display non-ASCII characters in the user interface and email servers to be able to process non-ASCII addresses correctly.
  • Technical standards to support EAI:

(i) SMTPUTF8: Allows email addresses with non-ASCII characters to be sent using Simple Mail Transfer Protocol (SMTP).

(ii) IDNA2008: Allows domain names with non-ASCII characters to be translated into ASCII-based Domain Name System (DNS) used by the internet.

Universal Acceptance

  • The basic idea behind Universal Acceptance is all domain names and all email addresses work in all software applications. It is a necessity for a truly inclusive internet to allow non-English speakers to access internet.
  • Universal Acceptance is a foundational requirement for a truly multilingual internet in which all internet users across the world can navigate entirely in local languages.
  • Essential for continued expansion of the internet and provides a gateway to the next billion internet users.
  • Enables the private sector, government and societies to better serve their communities through the use of an increasing number of new domains, including non-Latin based,

Universal Acceptance Day

  • Universal Acceptance Day is an awareness event celebrating the importance of ensuring that all domain names and email addresses can be used by all internet users.
  • In 2023, India was host of the First Global UA Day as the flag bearer to promote and promulgate Universal Acceptance for digital inclusion.
  • Global Universal Acceptance Day is scheduled to take place on March 28, 2024 in Belgrade, Serbia.

Universal Acceptance Guidelines

  • Universal Acceptance Guidelines are a set of best practices and recommendations for supporting the use of all domains names and email addresses, regardless of their script, language or format.
  • The guidelines for Universal Acceptance have been developed by Universal Acceptance Steering Group (UASG), a community led initiative that works to promote Universal Acceptance of all domain names and email addresses.
  • Universal Acceptance Steering Group (UASG) is a community led initiative that was formed in 2015 and funded by ICANN. It works to raise awareness of the importance of UA globally, provide free resources to organisations to help them become UA-ready and measure the progress of UA adoption.

NVIDIA’s Unveils Blackwell AI Chip

Context: NVIDIA has unveiled Blackwell AI Chip which are said to be the most powerful Artificial Intelligence enabled microchips in the world. 

What is GPU? Graphic Processing Unit

image 69

A Graphics Processing Unit (GPU) is a specialized circuit that accelerates image and video processing. GPUs have evolved into powerful parallel processors with applications beyond visual computing.

GPU Meaning and Usage: A GPU handles graphics and images in computers and phones, enhancing visuals in gaming and other applications.

Difference between CPU and GPU

  • Central Processing Unit (CPU): The brain of a computer system, consisting of the arithmetic logic unit (ALU) for calculations and the control unit (CU) for instruction sequencing and branching. It interacts with memory, input, and output components.
  • Graphics Processing Unit (GPU): A specialized processor for rendering images and graphics, particularly in computer games. GPUs are faster than CPUs and emphasize high throughput. They contain more ALU units than CPUs and share RAM with electronic equipment.
    • CPU focuses on low latency, while GPU focuses on high throughput.
    • CPU consumes more memory than GPU.
    • GPU is faster than CPU.
    • CPU has fewer, more powerful cores, while GPU has more, weaker cores.
    • CPU is suitable for serial instruction processing, while GPU is suitable for parallel instruction processing.

Technology behind GPUs

  • GPUs function by offloading parallelizable tasks from the CPU, leveraging data parallelism, APIs, shaders, and general-purpose computing capabilities.
  • Streaming Multiprocessors (SMs): SMs are the core processing units in GPUs, comprising multiple cores that operate concurrently to execute tasks in parallel.
  • Memory Hierarchy: GPUs have a dedicated memory hierarchy, including global memory, shared memory, and registers, to optimize data storage and access for enhanced performance.
  • Parallel Processing: GPUs excel in parallel processing, executing multiple operations simultaneously through the presence of multiple cores within SMs.

GPUs and Artificial Intelligence

GPUs have been called a rare-earth or even gold for Artificial Intelligence. This is due to the following reasons:

  • GPUs employ parallel processing.
  • GPU systems scale up to supercomputing heights.
  • The GPU software stack for AI is broad and deep.

GPUs perform technical calculations faster and with greater energy efficiency than CPUs.

Diverse Applications of GPUs

  • Gaming: GPUs provide realistic graphics and smooth animations for an immersive gaming experience.
  • Deep Learning and AI: GPUs accelerate matrix calculations essential for training and running deep neural networks.
  • Scientific Computing: GPUs expedite complex computations in simulations and calculations across various scientific domains.
  • Medical Imaging: GPU acceleration enables real-time rendering and analysis of medical imaging data.
  • Cryptocurrency Mining: GPUs efficiently handle the complex calculations required for validating cryptocurrency transactions.

GPU Applications Beyond Graphics

  • Data Science and Machine Learning: GPUs accelerate training and running complex machine learning models.
  • Cryptocurrency Mining: GPUs efficiently parallelize calculations for validating cryptocurrency transactions.
  • Computational Biology and Drug Discovery: GPUs accelerate simulations and analyses in biology and drug discovery.
  • Financial Modeling and Simulation: GPUs boost processing speed for complex financial models and simulations.
  • Autonomous Vehicles and Robotics: GPUs contribute to real-time object detection and decision-making in autonomous systems.

Challenges and Future Trends in GPU Technology

GPUs face challenges related to energy efficiency, ray tracing, quantum computing, and edge computing. Ongoing developments aim to address these challenges and shape the future of GPU technology.

The EU’s Artificial Intelligence Act

Context: European Union officials have reached a provisional deal on the world's first comprehensive laws to regulate the use of artificial intelligence. The European Parliament will vote on the AI Act proposals in early 2024, and the legislation is expected to take effect in 2025.

Major Highlights:

  • The European Parliament defines AI as software that can "for a given set of human-defined objectives, generate outputs such as content, predictions, recommendations or decisions influencing the environments they interact with".
  • The framework comprises safeguards on AI use within the European Union. These safeguards will include the framework through which consumers will be empowered to file complaints against violations, precise guardrails on AI adoption by law enforcement agencies and imposition of fine on violations. 
image 31

Classification of AI:

  • The Act’s central approach is the classification of AI techbased on the level of risk they pose to the “health and safety or fundamental rights” of a person. There are four risk categories in the Act — unacceptable risk, high risk, limited risk and minimal risk.
    • The Act prohibits using technologies in the unacceptable risk category with little exception. E.g., Deployment of facial recognition technology on a large scale, with a few exemptions for law enforcement.
    • The Act lays substantial focus on AI in the high-risk category, prescribing many pre-and post-market requirements for developers and users of such systems. E.g., use of AI tools for self-driving cars will be permitted, but it will be subject to certification.
    • AI systems in the limited (medium) and minimal risk category are allowed to be used with a few requirements like transparency obligations. E.g., generative AI chatbots, video games.  

Enforcement:

  • The EU will be able to monitor and sanction those who violate the law through a new body called the EU AI office
  • The EU AI office will have the power to slap a fine worth seven percent of a company's turnover or 35 million euros, whichever is larger.

Understanding Technology behind Generative Pre-Trained Transformer

Context: India will explore building large language models says Principal Scientific Advisor of India.

Natural Language Processing (NLP): 

  • NLP deals with giving computers the ability to understand text and spoken words in much the same way human beings can.
  • NLP combines human language with statistical, machine learning, and deep learning models. Together, these technologies enable computers to process human language in the form of text or voice data and to ‘understand’ its full meaning, complete with the speaker or writer’s intent and sentiment.
  • NLP drives computer programs that translate text from one language to another, respond to spoken commands, and summarize large volumes of text rapidly even in real-time. For example, voice-operated GPS systems, digital assistants, speech-to-text dictation software, customer service chatbots etc.

Understanding Large Language Model: 

  • Humans perceive the text as a collection of words. Sentences are sequences of words. Documents are sequences of chapters, sections, and paragraphs. But computers process one word or character at a time and provide an output once the entire input text has been consumed. 
  • LLM model works, but sometimes, it forgets what happened at the beginning of the sequence when the end is reached. So, computer scientists have found a transformer model to provide a better approach.

Transformer Model: 

 Process of Tokenisation 

  • To process a text input in a transformer model, the computer first tokenises the text input into a sequence of words.
  • These tokens are then encoded as numbers and converted into embeddings, which are vector-space representations of the tokens that preserve their meaning. 
  • Next, the encoder in the transformer transforms the embeddings of all the tokens into a context vector, which is like the essence of the entire text input. Using this vector, the transformer decoder generates output based on clues. 
  • Then, you can reuse the same decoder, but this time the clue will be the previously produced next word.
  • This process can be repeated to create an entire paragraph, starting from a leading sentence. This process is known as auto-regression.

In this model, the grammar of the output may not be correct as in reality, the transformer model doesn’t explicitly store grammar rules, instead, it acquires them implicitly through examples.

Large Language Model (LLM):

  • A large language model is a transformer model on a mass scale. 
  • It is so large that it usually cannot be run on a single computer. 
  • Such a large model is learned from a vast amount of text before it can remember the patterns and structures of language.
  • Because of this reason, it is naturally a service provided over API or a web interface. 

For example, the GPT-3 model which backs the ChatGPT service was trained on massive amounts of data from the internet including books, articles, websites etc. In this training it learned the statistical relationships between words, phrases, and sentences, allowing it to generate contextually relevant responses.

Other notable examples of this technology are Google's PaLM used in Bard, and Meta's LLaMa, as well as BLOOM, Ernie 3.0 Titan, and Anthropic's Claude 2.

LLM Use Cases:

  • Linguistic diversity and inclusion: India is a linguistically diverse country, with over 22 official languages and hundreds of dialects spoken. LLM can enable people to access information and services in their own language, reducing the digital divide and fostering greater inclusivity.
  • Economic growth and innovation: It can transform industries in India such as healthcare, education, and manufacturing. For example, LLMs can be used to develop new educational tools, improve consultation, and automate customer service tasks.
  • Personalisation and training: It can generate training data, modules etc. and develop educational and similar tools that are personalised to each person’s needs making education and delivery of services effective.
  • Use in Research and Development (R&D): LLMs can be used to analyse large datasets of scientific data to identify patterns and trends that would be difficult or impossible for humans to find on their own. E.g., IIT-D is using LLM to find Malaria drugs.
  • Use in Intelligence: By analysing large social media datasets, satellite imagery, financial transaction data and human intelligence data and finding patterns through LLM intelligence agencies can identify potential threats, extremist propaganda etc. E.g., the US National Security Agency is using LLMs to analyse large amounts of social media data to identify potential terrorist threats. Hence, LLMs can be used to develop new tools for disseminating Early warnings, cybersecurity, social media monitoring etc. 

Concerns about the govt.’s fact check unit

Context: Bombay High court has recently reserved its judgement on petitions challenging constitutionality of the Fact Check Unit established under the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2023 (IT Rules) in Kunal Kamra v. Union of India.

Background:

  • In 2023, MEiTY promulgated the 2023 rules amending IT rules 2021, allowing the government to constitute a Fact Checking Unit (FCU) under IT (Intermediary Guidelines and Digital Media Ethics Code), Amendment 2023.
  • New rules require Social Media intermediaries to censor or modify the content relating to “business of Central government” if the FCU directs them to do so.
  • FCU will have power to decide whether the content on social media is fake, false, or misleading.
  • Once content is identified as fake or misleading, intermediaries must take action or lose “safe harbor” protection under I.T. Act, 2000.

Concerns against Fact Check Unit

  • Change is not in spirit with Section 79A of IT Act, 2000 which mandates providing details with reasons for such takedown orders given to IT intermediaries.
  • As there is no right to appeal or judicial oversight over orders given by the Fact Check Unit, such powers can be misused to prevent questioning or scrutiny by media organizations.
  • May create a “chilling effect” on the right to speech and expression on digital platforms which is against the spirit of Article 19(1)(a) as it gives unfettered powers to FCU.
  • Violative of Right to Equality under Article 14 as it discriminates between false news about the government and other false news.
  • Violates separation of powers as these orders have a judicial overtone but exercised by an executive unit i.e., the Fact Check Unit.
  • There is a moral argument, government cannot be the sole arbiter of what constitutes fake news.
  • Broad scope open to misuse as the FCU can verify the veracity and take down any post on IT intermediaries which relates to “any business of central government”, which is very broad and not defined.
  • No clear parameters or guidelines have been given on the grounds on which FCU flags an information as Fake or ambiguous.
  • Concerns as to what part of information would be fake, whether the source is fake, or the information itself is fake.
  • Rules do not offer protection to fair criticism of government.
  • PIB is already doing fact checking efficiently, so why need a new mechanism.
  • Ambiguity with respect to the well settled test of reasonableness under Article 19(2) with respect to the boundary separating fake news.

Other ethical issues

  • Misleading for one may not be for the other, such terms are subjective, and open to various interpretations.
  • In editorial content such as Economy, there can be variation in data as per government and other researchers, such data cannot be called as fake data.
  • There are no provisions in Rules that provide an opportunity to justify or defend the flagged content, violating principles of natural justice.

Rationale of Government for establishing FCU

  • FCU’s notices are merely advisory in nature.
  • FCU will only notify intermediaries that the content is fake, or misleading. Intermediaries can choose to take it down or leave with disclaimer.
  • In case the user is aggrieved by the intermediary’s decision, they can avail remedy before the court of law.

Conclusion & way forward

Law has to be tested not only on its intentions, but also if its consequences are constitutional.

In Bennett Coleman & Co. vs Union of India, S.C. has held that freedom of speech and of press is the Arc of the Covenant of democracy because public criticism is essential to the working of its institutions, which must be respected.·   Moving ahead with FCU, the government must ensure that the constitution of FCU restricts to the test of reasonableness under Article 19(2) and guidelines given in Shreya Singhal vs Union of India (2015).

BlueWalker 3 Satellite

Context: An international team of scientists has published a paper that outlines the impact of the prototype BlueWalker 3 satellite on astronomy.

About BlueWalker 3:

  • It is a prototype satellite, a part of a planned constellation of over a hundred similar satellites intended to deliver mobile or broadband services anywhere in the world.
    • The approximately 1.5-ton satellite will deploy a 10-meter diameter phased array antenna, comprising numerous identical sub-antenna modules with a total area of 64 square meters. These modules will connect directly to standard mobile phones.
  • It was launched into low-Earth orbit in 2022 by AST SpaceMobile, a U.S.-based company.
  • It is considered the largest commercial antenna system ever deployed in low-Earth orbit.
  • The satellite is among the brightest objects in the sky reaching a peak comparable to that of Procyon and Achernar, two of the brightest stars in the night sky.
  • Its remarkable brightness results from a massive phased-array antenna, making it appear like a giant mirror reflecting sunlight from Earth’s perspective. 
BlueWalker 3 Satellite image
Image Source: ast-science.com

Concerns:

  • Bluewalker 3’s large size and bright reflective surfaces could interfere with astronomical observations, as its light could be mistaken for stars or interfere with the ability to detect dimmer objects.
  • The satellite’s large size could block out a portion of the night sky, making it difficult for astronomers to observe certain objects.
  • It actively transmits at radio frequencies that are close to bands reserved for radio astronomy, which may hamper radio telescope observations