Artificial Intelligence

1. Foundational AI Concepts

Summary of AI Foundations


Define Artificial Intelligence, its current status, and key technological enablers.	- Artificial Intelligence (AI): The ability of machines to simulate human cognitive functions, such as perceiving, natural language understanding, learning, reasoning, and problem-solving. Current Status: Primarily Narrow AI (Weak AI), which is intelligent only within specific contexts. Key Enablers: Massive increases in computational power (e.g., TPUs/GPUs) and the availability of Big Data.
Differentiate between Narrow AI, AGI, and Superintelligence.	- Narrow AI (Weak AI): Designed for highly specific tasks (e.g., AlphaGo, Siri, recommendation engines). Cannot generalize outside its programmed domain. - Artificial General Intelligence (AGI): Human-level intelligence across a wide range of cognitive tasks. Currently a technological frontier, not yet achieved. - Superintelligence: A theoretical future state where machine intelligence far surpasses human capabilities across all domains. Represents a long-term existential risk.

Define Artificial Intelligence, its current status, and key technological enablers.

- Artificial Intelligence (AI): The ability of machines to simulate human cognitive functions, such as perceiving, natural language understanding, learning, reasoning, and problem-solving.
Current Status: Primarily Narrow AI (Weak AI), which is intelligent only within specific contexts.
Key Enablers: Massive increases in computational power (e.g., TPUs/GPUs) and the availability of Big Data.

Differentiate between Narrow AI, AGI, and Superintelligence.

- Narrow AI (Weak AI): Designed for highly specific tasks (e.g., AlphaGo, Siri, recommendation engines). Cannot generalize outside its programmed domain. - Artificial General Intelligence (AGI): Human-level intelligence across a wide range of cognitive tasks. Currently a technological frontier, not yet achieved. - Superintelligence: A theoretical future state where machine intelligence far surpasses human capabilities across all domains. Represents a long-term existential risk.

AI Approaches & Key Technologies


Compare Rule-based Systems and Machine Learning (ML) approaches.	- Rule-based Systems: Rely on human-encoded rules and logic (e.g., Deep Blue chess computer). Transparent and predictable, but fail in complex, dynamic, or uncertain scenarios. Machine Learning (ML): Uses statistical algorithms to learn patterns and make predictions directly from large datasets. Generative AI: A sub-field of ML capable of creating original content (text, images, code, video) (e.g., ChatGPT, Gemini, DALL-E).
Detail the key technology areas within the AI landscape.	- Computer Vision: Interpreting and analyzing visual data (e.g., facial recognition, autonomous driving). - Natural Language Processing (NLP): Comprehending and generating human language (e.g., machine translation, virtual assistants). - Speech Recognition: Converting spoken language into structured text. - Robotics: Combining AI with physical systems for autonomous movement and action. - Predictive Analytics: Forecasting future trends based on historical data patterns (e.g., weather, fraud detection).

Compare Rule-based Systems and Machine Learning (ML) approaches.

- Rule-based Systems: Rely on human-encoded rules and logic (e.g., Deep Blue chess computer). Transparent and predictable, but fail in complex, dynamic, or uncertain scenarios.
Machine Learning (ML): Uses statistical algorithms to learn patterns and make predictions directly from large datasets.
Generative AI: A sub-field of ML capable of creating original content (text, images, code, video) (e.g., ChatGPT, Gemini, DALL-E).

Detail the key technology areas within the AI landscape.

- Computer Vision: Interpreting and analyzing visual data (e.g., facial recognition, autonomous driving). - Natural Language Processing (NLP): Comprehending and generating human language (e.g., machine translation, virtual assistants). - Speech Recognition: Converting spoken language into structured text. - Robotics: Combining AI with physical systems for autonomous movement and action. - Predictive Analytics: Forecasting future trends based on historical data patterns (e.g., weather, fraud detection).

2. Sectoral Applications & Major Concerns

Applications (Sectors)


Outline high-yield AI applications in Healthcare, Agriculture, and Defence.	- Healthcare: Early cancer detection (radiology analysis), computer-aided drug discovery, and precision robotic surgery. - Agriculture: Smart crop planning, satellite/drone-based pathogen detection, and predictive yield modeling. - Defence: Unmanned autonomous combat vehicles, automated border patrolling, and intelligence analysis.
Outline high-yield AI applications in Judiciary, Transport, and Policing.	- Judiciary: Automated legal research, multi-lingual case translation, and historical outcome prediction. - Transport: Self-driving vehicles, traffic management, and logistics route optimization. - Policing: Facial recognition databases and predictive policing to map crime hotspots.

Core AI Concerns


Explain AI Hallucinations and the demographic error rates in facial recognition.	- Hallucinations: A phenomenon where Generative AI models generate false, nonsensical, or incorrect information but present it with high confidence as fact. Demographic Bias: Facial recognition models exhibit significantly higher error rates when identifying female and non-white demographics due to skewed training data.
Define the 'Black Box' problem in AI systems.	- Black Box Problem: The lack of explainability (difficulty tracing how deep learning models arrive at specific decisions) and transparency (proprietary datasets and algorithms kept hidden from the public).
How does AI training clash with Privacy and Intellectual Property (IPR) laws?	- Privacy: Conflict between the data minimization principle of privacy laws (e.g., DPDP Act) and AI's massive data consumption needs. IPR: Fair use debate regarding scraping copyrighted original works to train LLMs, and legal ambiguity over whether AI-generated content can be copyrighted.
What is the projected impact of AI on labor and employment?	- Labor Disruption: Automation of non-routine, creative, and analytical tasks. It threatens clerical, writing, and coding jobs, while simultaneously creating new roles in AI safety, prompt engineering, and data curation.

3. Regulation & Global Frameworks

AI Regulation & Indian Initiatives


Contrast the global AI regulatory models of the EU, USA, and China.	- European Union (EU): Implemented a strict, risk-based classification framework under the EU AI Act (banning unacceptable-risk systems, regulating high-risk models). - USA: Relies on voluntary standards, safety guidelines, and developer self-reporting mandated by the Oct 2023 Executive Order. - China: Targets specific algorithmic systems, requiring strict registration and compliance with state guidelines for generative AI content.
Detail India's national regulatory and policy initiatives for AI.	- NITI Aayog: Released the "Responsible AI for All" strategy and framework (2021). Digital India Act (Proposed): Intended to introduce legal guidelines regulating high-risk AI applications and deepfakes. IndiaAI Mission: India's flagship programme funded with ₹10,372 crore to build compute capacity, support startups, and develop public data platforms.

Contrast the global AI regulatory models of the EU, USA, and China.

- European Union (EU): Implemented a strict, risk-based classification framework under the EU AI Act (banning unacceptable-risk systems, regulating high-risk models). - USA: Relies on voluntary standards, safety guidelines, and developer self-reporting mandated by the Oct 2023 Executive Order. - China: Targets specific algorithmic systems, requiring strict registration and compliance with state guidelines for generative AI content.

Detail India's national regulatory and policy initiatives for AI.

- NITI Aayog: Released the "Responsible AI for All" strategy and framework (2021).
Digital India Act (Proposed): Intended to introduce legal guidelines regulating high-risk AI applications and deepfakes.
IndiaAI Mission: India's flagship programme funded with ₹10,372 crore to build compute capacity, support startups, and develop public data platforms.

4. Frontier Developments (2025-2026 Update)

Indigenous Innovations & Models


What is Bharat Gen, its developer, and its implementation project?	- Bharat Gen: India's first indigenous Multimodal Large Language Model (LLM) designed to support all 22 scheduled Indian languages. Developer: Led by IIT Bombay under the National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS) to make generative AI accessible locally.
Explain Agentic AI and C2S-SCALE.	- Agentic AI: Systems capable of acting autonomously, planning, executing multi-step tasks, and making decisions without constant human prompts (e.g., Kruti). C2S-SCALE: A specialized space-sector initiative (Chip to Startup - Space Capability for AI and Landing Excellence) applying edge AI inside satellite systems.
What are AlphaGenome and AlphaFold2?	- AlphaGenome: AI engine specialized in genomic sequencing and DNA variant effect analysis. AlphaFold2: Deep learning model that predicts 3D protein structures with atomic accuracy from amino acid sequences (highly celebrated in Nobel Prize contexts).
Define Suncatcher, Ironwood, and the Willow processor.	- Suncatcher & Ironwood: Google's next-generation TPU (Tensor Processing Unit) accelerator infrastructures optimizing large-scale LLM training. Willow: Google's state-of-the-art quantum processor, demonstrating a ~13,000x computational speedup over classical supercomputers.

AI Economic Paradoxes & Deepfakes


Differentiate between Jevons Paradox and Solow Paradox in AI economics.	- Jevons Paradox: When technological efficiency reduces the cost of a resource (e.g., AI compute), it actually increases the overall consumption of that resource rather than reducing it. Solow Paradox: The empirical observation that massive investments in information technology and AI have not translated into measurable gains in economy-wide productivity statistics.
Explain the neural networks that power deepfakes, and India's regulatory safeguards.	Generative Adversarial Networks (GANs): Comprises a Generator (creates synthetic data) and an Adversary/Discriminator (tests and refines the realism) in a feedback loop. Autoencoders are also used. Indian Legal Safeguards: - IT Rules 2021 (Amendments): Mandate labeling of Synthetic Generated Information (SGI) and prompt takedowns. - DPDP Act 2023: Restricts unauthorized personal data scraping to generate deepfakes.

Differentiate between Jevons Paradox and Solow Paradox in AI economics.

- Jevons Paradox: When technological efficiency reduces the cost of a resource (e.g., AI compute), it actually increases the overall consumption of that resource rather than reducing it.
Solow Paradox: The empirical observation that massive investments in information technology and AI have not translated into measurable gains in economy-wide productivity statistics.

Explain the neural networks that power deepfakes, and India's regulatory safeguards.

Generative Adversarial Networks (GANs): Comprises a Generator (creates synthetic data) and an Adversary/Discriminator (tests and refines the realism) in a feedback loop. Autoencoders are also used.
Indian Legal Safeguards: - IT Rules 2021 (Amendments): Mandate labeling of Synthetic Generated Information (SGI) and prompt takedowns. - DPDP Act 2023: Restricts unauthorized personal data scraping to generate deepfakes.

🎯 UPSC Prelims 2026 Questions & Explanations


can be used to eliminate options c and d. Which of the following st...	- Question 68: can be used to eliminate options c and d. Which of the following statements with regard to stealth technology is/are correct? Correct Answer: (C) Key Concepts & Explanation: and self-attention to analyze complex data sets. 2. MT-NLG is the world's largest language model developed by NVIDIA and Microsoft. 3. Large language models are trained using supervised learning. Which of the statements given above are correct? (a) 1 and 2 only (b) 2 and 3 only (c) 1 and 3 only (d) 1,2 and 3 Stealth technology is intended to reduce the detectability of aircraft, ships, and other military platforms by radar, infrared systems, and other surveillance methods. One of its primary objectives is minimizing the radar cross-section (RCS) through specialized shaping and the use of Radar Absorbing Materials (RAM). Stealth technology in defence development and innovation M F https://indianexpress.com/article/explained/this-word-means-fifth-generation-fighter-10034632/ EN 1. Stealth objects have a very small radar cross-section and are coated with Radar Absorbing Material. 2. Stealth objects can be detected using specific frequencies. 3. Stealth objects are coated with metamaterialstoincrease the scattering of electromagnetic radiation. Year Asked: Prelims 2026
Consider the following statements with respect to the AI Impact Sum...	- Question 90: Consider the following statements with respect to the AI Impact Summit, 2026 held in New Delhi: 1. The Summit's intellectual framework was based on three foundational Sutras: People, Planning, and Progress. 2. The Preamble of the Summit stresses Democratising AI Resources, which acknowledges the Charter for Democratic Diffusion of AI as a binding framework to support locally relevant innovation and strengthen resilient AI ecosystems while respecting national laws. 3. The New Delhi Declaration on AI Impact was structured around seven Access Chakras (Pillars), which included Correct Answer: (D) Key Concepts & Explanation: The AI Impact Summit 2026, held in New Delhi, focused on shaping the future direction of artificial intelligence through an inclusive and development-oriented approach. The summit adopted an intellectual framework based on three guiding Sutras — People, Planet, and Prosperity. It also emphasized the democratization of AI resources and introduced the New Delhi Declaration on AI Impact, which was organized around seven thematic Chakras or pillars addressing key areas of AI development and governance. The summit framework was built around the three guiding principles of People, Planet, and Prosperity rather than People, Planning, and Progress. Hence, statement 1 is not correct. While the preamble emphasizes the democratization of AI resources, the Charter for Democratic Diffusion of AI was envisaged as a voluntary and non-binding framework rather than a legally binding arrangement. Hence, statement is incorrect. From the recent Feb 2026 summit (also mentioned in the stem) M CA https://www.pi.b.gov.in/PressR.eleasePage.aspx?PRID=2225069®=3&lang=2 EN PT365 2026 Update as.in/cd 2: Article 5.3. AI IMPACT 2026-SUMM102-T Monthence-y CurrenetAffairsFebruary2026:Article7.1.AIImpactSummit-tNewsToday(17 Feb2026):'India' Year Asked: Prelims 2026

can be used to eliminate options c and d. Which of the following st...

- Question 68: can be used to eliminate options c and d.
Which of the following statements with regard to stealth technology is/are correct?

Correct Answer: (C)

Key Concepts & Explanation:
and self-attention to
analyze complex
data sets.
2. MT-NLG is the world's largest language model developed by NVIDIA and Microsoft.
3. Large language models are trained using supervised learning.
Which of the statements given above are correct?
(a) 1 and 2 only
(b) 2 and 3 only
(c) 1 and 3 only
(d) 1,2 and 3
Stealth technology is intended to reduce the detectability of aircraft, ships, and other military platforms by radar, infrared systems, and other surveillance methods. One of its primary objectives is minimizing the radar cross-section (RCS) through specialized shaping and the use of Radar Absorbing Materials (RAM).
Stealth technology in defence development and innovation
M
F https://indianexpress.com/article/explained/this-word-means-fifth-generation-fighter-10034632/
EN
1. Stealth objects have a very small radar cross-section and are coated with Radar Absorbing Material.
2. Stealth objects can be detected using specific frequencies.
3. Stealth objects are coated with metamaterialstoincrease the scattering of electromagnetic radiation.

Year Asked: Prelims 2026

Consider the following statements with respect to the AI Impact Sum...

- Question 90: Consider the following statements with respect to the AI Impact Summit, 2026 held in New Delhi:
1. The Summit's intellectual framework was based on three foundational Sutras:
People, Planning, and Progress.
2. The Preamble of the Summit stresses
Democratising AI Resources, which acknowledges the Charter for Democratic Diffusion of AI as a binding framework to support locally relevant innovation and strengthen resilient AI ecosystems while respecting national laws.
3. The New Delhi Declaration on AI Impact was structured around seven Access Chakras (Pillars), which included

Correct Answer: (D)

Key Concepts & Explanation:
The AI Impact Summit 2026, held in New Delhi, focused on shaping the future direction of artificial intelligence through an inclusive and development-oriented approach. The summit adopted an intellectual framework based on three guiding Sutras — People, Planet, and Prosperity. It also emphasized the democratization of AI resources and introduced the New Delhi Declaration on AI Impact, which was organized around seven thematic Chakras or pillars addressing key areas of AI development and governance.
The summit framework was built around the three guiding principles of People, Planet, and Prosperity rather than People, Planning, and Progress. Hence, statement 1 is not correct. While the preamble emphasizes the democratization of AI resources, the Charter for Democratic Diffusion of AI was envisaged as a voluntary and non-binding framework rather than a legally binding arrangement. Hence, statement is incorrect.
From the recent Feb 2026 summit (also mentioned in the stem)
M
CA
https://www.pi.b.gov.in/PressR.eleasePage.aspx?PRID=2225069®=3&lang=2
EN
PT365 2026 Update as.in/cd 2: Article 5.3. AI IMPACT 2026-SUMM102-T Monthence-y CurrenetAffairsFebruary2026:Article7.1.AIImpactSummit-tNewsToday(17 Feb2026):'India'

Year Asked: Prelims 2026

Artificial Intelligence ​

1. Foundational AI Concepts ​

2. Sectoral Applications & Major Concerns ​

3. Regulation & Global Frameworks ​

4. Frontier Developments (2025-2026 Update) ​