#FactCheck: Old Salman Khan Video From 2019 Misleadingly Shared as Eid Post

Artificial intelligence is revolutionizing industries such as healthcare to finance to influence the decisions that touch the lives of millions daily. However, there is a hidden danger associated with this power: unfair results of AI systems, reinforcement of social inequalities, and distrust of technology. One of the main causes of this issue is training data bias, which appears when the examples on which an AI model is trained are not representative or skewed. To deal with it successfully, this needs a combination of statistical methods, algorithmic design that is mindful of fairness, and robust governance over the AI lifecycle. This article discusses the origin of bias, the ways to reduce it, and the unique position of fairness-conscious algorithms.

‍

Why Bias in Training Data Matters

‍

The bias in AI occurs when the models mirror and reproduce the trends of inequality in the training data. When a dataset has a biased representation of a demographic group or includes historical biases, the model will be trained to make decisions in ways that will harm the group. This is a fact that has a practical implication: prejudiced AI may cause discrimination during the recruitment of employees, lending, and evaluation of criminal risks, as well as various other spheres of social life, thus compromising justice and equity. These problems are not only technical in nature but also require moral principles and a system of governance (E&ICTA).

Bias is not uniform. It may be based on the data itself, the algorithm design, or even the lack of diversity among developers. The bias in data occurs when data does not represent the real world. Algorithm bias may arise when design decisions inadvertently put one group at an unfair advantage over another. Both the interpretation of the model and data collection may be affected by human bias. (MDPI)

‍

Statistical Principles for Reducing Training Data Bias 

‍

Statistical principles are at the core of bias mitigation and they redefine the data-model interaction. These approaches are focused on data preparation, training process adjustment, and model output corrections in such a way that the notion of fairness becomes a quantifiable goal.

‍

Balancing Data Through Re-Sampling and Re-Weighting

‍

Among the aforementioned methods, a fair representation of all the relevant groups in the dataset is one way. This can be achieved by oversampling underrepresented groups and undersampling overrepresented groups. Oversampling gives greater weight to minority examples, whereas re-weighting gives greater weight to under-represented data points in training. The methods minimize the tendency of models to fit to salient patterns and improve coverage among vulnerable groups. (GeeksforGeeks)

‍

Feature Engineering and Data Transformation

‍

The other statistical technique is to convert data characteristics in such a way that sensitive characteristics have a lesser impact on the results. In one example, fair representation learning adjusts the data representation to discourage bias during the untraining of the model. The disparate impact remover adjust technique performs the adjustment of features of the model in such a way that the impact of sensitive features is reduced during learning. (GeeksforGeeks)

‍

Measuring Fairness With Metrics

‍

Statistical fairness measures are used to measure the effectiveness of a model in groups. 

‍

Fairness-Aware Algorithms Explained

‍

Fair algorithms do not simply detect bias. They incorporate fairness goals in model construction and run in three phases including pre-processing, in-processing, and post-processing.

‍

Pre-Processing Techniques

‍

Fairness-aware pre-processing deals with bias prior to the model consuming the information. This involves the following ways:

‍

• Rebalancing training data through sampling and re-weighting training data to address sample imbalances.
• Data augmentation to generate examples of underrepresented groups.
• Feature transformation removes or downplays the impact of sensitive attributes prior to the commencement of training. (IJMRSET)

‍

These methods can be used to guarantee that the model is trained on more balanced data and to reduce the chances of bias transfer between historical data.

‍

In-Processing Techniques

‍

The in-processing techniques alter the learning algorithm. These include:

‍

• Fairness constraints that penalize the model for making biased predictions during training.
• Adversarial debiasing, where a second model is used to ensure that sensitive attributes are not predicted by the learned representations.
• Fair representation learning that modifies internal model representations in favor of 

‍

Post-Processing Techniques

‍

Fairness may be enhanced after training by changing the model outputs. These strategies comprise:

‍

• Threshold adjustments to various groups to meet conditions of fairness, like equalized odds.
• Calibration techniques such that the estimated probabilities are fair indicators of the actual probabilities in groups. (GeeksforGeeks)

‍

Challenges 

‍

Mitigating bias is complex. The statistical bias minimization may at times come at the cost of the model accuracy, and there is a conflict between predictive performance and fairness. The definition of fairness itself is potentially a difficult task because various applications of fairness require various criteria, and various criteria can be conflicting. (MDPI)

Gaining varied and representative data is also a challenge that is experienced because of privacy issues, incomplete records, and a lack of resources. The auditing and reporting done on a continuous basis are needed so that mitigation processes are up to date, as models are continually updated. (E&ICTA)

‍

Why Fairness-Aware Development Matters

‍

The outcomes of the unfair treatment of some groups by AI systems are far-reaching. Discriminatory software in recruitment may support inequality in the workplace. Subjective credit rating may deprive deserving people of opportunities. Unbiased medical forecasts might result in the flawed allocation of medical resources. In both cases, prejudice contravenes the credibility and clouds the greater prospect of AI. (E&ICTA)

Algorithms that are fair and statistical mitigation plans provide a way to create not only powerful AI but also fair and trustworthy AI. They admit that the results of AI systems are social tools whose effects extend across society. Responsible development will necessitate sustained fairness quantification, model adjustment, and upholding human control.

‍

Conclusion 

‍

AI bias is not a technical malfunction. It is a mirror of real-world disparities in data and exaggerated by models. Statistical rigor, wise algorithm design, and readiness to address the trade-offs between fairness and performance are required to reduce training data bias. Fairness-conscious algorithms (which can be implemented in pre-processing, in-processing, or post-processing) are useful in delivering more fair results. As AI is taking part in the most crucial decisions, it is necessary to consider fairness at the beginning to have a system that serves the population in a responsible and fair manner.

‍

References

‍

• Understanding Bias in Artificial Intelligence: Challenges, Impacts, and Mitigation Strategies: E&ICTA, IITK
• Bias and Fairness in Artificial Intelligence: Methods and Mitigation Strategies: JRPS Shodh Sagar 
• Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies: MDPI 
• Ensuring Fairness in Machine Learning Algorithms: GeeksforGeeks
  Bias and Fairness in Machine Learning Models: A Critical Examination of Ethical Implications: IJMRSET 
• Bias in AI Models: Origins, Impact, and Mitigation Strategies: Preprints
• Bias in Artificial Intelligence and Mitigation Strategies: TCS
• Survey on Machine Learning Biases and Mitigation Techniques: MDPI

‍

What Is a VPN and its Significance 

A Virtual Private Network (VPN) creates a secure and reliable network connection between a device and the internet. It hides your IP address by rerouting it through the VPN’s host servers. For example, if you connect to a US server, you appear to be browsing from the US, even if you’re in India. It also encrypts the data being transferred in real-time so that it is not decipherable by third parties such as ad companies, the government, cyber criminals, or others.

All online activity leaves a digital footprint that is tracked for data collection, and surveillance, increasingly jeopardizing user privacy. VPNs are thus a powerful tool for enhancing the privacy and security of users, businesses, governments and critical sectors. They also help protect users on public Wi-Fi networks ( for example, at airports and hotels), journalists, activists and whistleblowers, remote workers and businesses, citizens in high-surveillance states, and researchers by affording them a degree of anonymity. 

What VPNs Do and Don’t 

‍

•  What VPNs Can Do:
  
  • Mask your IP address to enhance privacy.
  • Encrypt data to protect against hackers, especially on public Wi-Fi.
  • Bypass geo-restrictions (e.g., access streaming content blocked in India).
•  What VPNs Cannot Do:
  
  • Make you completely anonymous and protect your identity (websites can still track you via cookies, browser fingerprinting, etc.).
  • Protect against malware or phishing.
  • Prevent law enforcement from tracing you if they have access to VPN logs.
  • Free VPNs usually even share logs with third parties. 

VPNs in the Context of India’s Privacy Policy Landscape 

In April 2022, CERT-In (Computer Emergency Response Team- India) released Directions under Section 70B (6) of the Information Technology (“IT”) Act, 2000, mandating VPN service providers to store customer data such as “validated names of subscribers/customers hiring the services, period of hire including dates, IPs allotted to / being used by the members, email address and IP address and time stamp used at the time of registration/onboarding, the purpose for hiring services, validated address and contact numbers, and the ownership pattern of the subscribers/customers hiring services” collected as part of their KYC (Know Your Customer) requirements, for a period of five years, even after the subscription has been cancelled. While this directive was issued to aid with cybersecurity investigations, it undermines the core purpose of VPNs- anonymity and privacy. It also gave operators very little time to carry out compliance measures. 

Following this, operators such as NordVPN, ExpressVPN, ProtonVPN, and others pulled their physical servers out of India, and now use virtual servers hosted abroad (e.g., Singapore) with Indian IP addresses. While the CERT-In Directions have extra-territorial applicability, virtual servers are able to bypass them since they physically operate from a foreign jurisdiction. This means that they are effectively not liable to provide user information to Indian investigative agencies, beating the whole purpose of the directive. To counter this, the Indian government could potentially block non-compliant VPN services in the future. Further, there are concerns about overreach since the Directions are unclear about how long CERT-In can retain the data it acquires from VPN operators, how it will be used and safeguarded, and the procedure of holding VPN operators responsible for compliance. 

Conclusion: The Need for a Privacy-Conscious Framework

The CERT-In Directions reflect a governance model which, by prioritizing security over privacy, compromises on safeguards like independent oversight or judicial review to balance the two. The policy design renders a lose-lose situation because virtual VPN services are still available, while the government loses oversight. If anything, this can make it harder for the government to track suspicious activity.  It also violates the principle of proportionality established in the landmark privacy judgment, Puttaswamy v. Union of India (II) by giving government agencies the power to collect excessive VPN data on any user. These issues underscore the need for a national-level, privacy-conscious cybersecurity framework that informs other policies on data protection and cybercrime investigations. In the meantime, users who use VPNs are advised to choose reputable providers, ensure strong encryption, and follow best practices to maintain online privacy and security.

References 

• https://www.kaspersky.com/resource-center/definitions/what-is-a-vpn 
• https://internetfreedom.in/top-secret-one-year-on-cert-in-refuses-to-reveal-information-about-compliance-notices-issued-under-its-2022-directions-on-cybersecurity/#:~:text=tl;dr,under%20this%20new%20regulatory%20mandate. 
• https://www.wired.com/story/vpn-firms-flee-india-data-collection-law/#:~:text=Starting%20today%2C%20the%20Indian%20Computer,years%2C%20even%20after%20they%20have

‍

Introduction

On the precipice of a new domain of existence, the metaverse emerges as a digital cosmos, an expanse where the horizon is not sky, but a limitless scope for innovation and imagination. It is a sophisticated fabric woven from the threads of social interaction, leisure, and an accelerated pace of technological progression. This new reality, a virtual landscape stretching beyond the mundane encumbrances of terrestrial life, heralds an evolutionary leap where the laws of physics yield to the boundless potential inherent in our creativity. Yet, the dawn of such a frontier does not escape the spectre of an age-old adversary—financial crime—the shadow that grows in tandem with newfound opportunity, seeping into the metaverse, where crypto-assets are no longer just an alternative but the currency du jour, dazzling beacons for both legitimate pioneers and shades of illicit intent.

The metaverse, by virtue of its design, is a canvas for the digital repaint of society—a three-dimensional realm where the lines between immersive experiences and entertainment blur, intertwining with surreal intimacy within this virtual microcosm. Donning headsets like armor against the banal, individuals become avatars; digital proxies that acquire the ability to move, speak, and perform an array of actions with an ease unattainable in the physical world. Within this alternative reality, users navigate digital topographies, with experiences ranging from shopping in pixelated arcades to collaborating in virtual offices; from witnessing concerts that defy sensory limitations to constructing abodes and palaces from mere codes and clicks—an act of creation no longer beholden to physicality but to the breadth of one's ingenuity.

The Crypto Assets

The lifeblood of this virtual economy pulsates through crypto-assets. These digital tokens represent value or rights held on distributed ledgers—a technology like blockchain, which serves as both a vault and a transparent tapestry, chronicling the pathways of each digital asset. To hop onto the carousel of this economy requires a digital wallet—a storeroom and a gateway for acquisition and trade of these virtual valuables. Cryptocurrencies, with NFTs—Non-fungible Tokens—have accelerated from obscure digital curios to precious artifacts. According to blockchain analytics firm Elliptic, an astonishing figure surpassing US$100 million in NFTs were usurped between July 2021 and July 2022. This rampant heist underlines their captivating allure for virtual certificates. Empowers do not just capture art, music, and gaming, but embody their very soul.

Yet, as the metaverse burgeons, so does the complexity and diversity of financial transgressions. From phishing to sophisticated fraud schemes, criminals craft insidious simulacrums of legitimate havens, aiming to drain the crypto-assets of the unwary. In the preceding year, a daunting figure rose to prominence—the vanishing of US$14 billion worth of crypto-assets, lost to the abyss of deception and duplicity. Hence, social engineering emerges from the shadows, a sort of digital chicanery that preys not upon weaknesses of the system, but upon the psychological vulnerabilities of its users—scammers adorned in the guise of authenticity, extracting trust and assets with Machiavellian precision.

The New Wave of Fincrimes 

Extending their tentacles further, perpetrators of cybercrime exploit code vulnerabilities, engage in wash trading, obscuring the trails of money laundering, meander through sanctions evasion, and even dare to fund activities that send ripples of terror across the physical and virtual divide. The intricacies of smart contracts and the decentralized nature of these worlds, designed to be bastions of innovation, morph into paths paved for misuse and exploitation. The openness of blockchain transactions, the transparency that should act as a deterrent, becomes a paradox, a double-edged sword for the law enforcement agencies tasked with delineating the networks of faceless adversaries.

Addressing financial crime in the metaverse is Herculean labour, requiring an orchestra of efforts—harmonious, synchronised—from individual users to mammoth corporations, from astute policymakers to vigilant law enforcement bodies. Users must furnish themselves with critical awareness, fortifying their minds against the siren calls that beckon impetuous decisions, spurred by the anxiety of falling behind. Enterprises, the architects and custodians of this digital realm, are impelled to collaborate with security specialists, to probe their constructs for weak seams, and to reinforce their bulwarks against the sieges of cyber onslaughts. Policymakers venture onto the tightrope walk, balancing the impetus for innovation against the gravitas of robust safeguards—a conundrum played out on the global stage, as epitomised by the European Union's strides to forge cohesive frameworks to safeguard this new vessel of human endeavour.

The Austrian Example 

Consider the case of Austria, where the tapestry of laws entwining crypto-assets spans a gamut of criminal offences, from data breaches to the complex webs of money laundering and the financing of dark enterprises. Users and corporations alike must become cartographers of local legislation, charting their ventures and vigilances within the volatile seas of the metaverse.

Upon the sands of this virtual frontier, we must not forget: that the metaverse is more than a hive of bits and bandwidth. It crystallises our collective dreams, echoes our unspoken fears, and reflects the range of our ambitions and failings. It stands as a citadel where the ever-evolving quest for progress should never stray from the compass of ethical pursuit. The cross-pollination of best practices, and the solidarity of international collaboration, are not simply tactics—they are imperatives engraved with the moral codes of stewardship, guiding us to preserve the unblemished spirit of the metaverse.

Conclusion

The clarion call of the metaverse invites us to venture into its boundless expanse, to savour its gifts of connection and innovation. Yet, on this odyssey through the pixelated constellations, we harness vigilance as our star chart, mindful of the mirage of morality that can obfuscate and lead astray. In our collective pursuit to curtail financial crime, we deploy our most formidable resource—our unity—conjuring a bastion for human ingenuity and integrity. In this, we ensure that the metaverse remains a beacon of awe, safeguarded against the shadows of transgression, and celebrated as a testament to our shared aspiration to venture beyond the realm of the possible, into the extraordinary.

References 

• https://www.wolftheiss.com/insights/financial-crime-in-the-metaverse-is-real/ 
• https://gnet-research.org/2023/08/16/meta-terror-the-threats-and-challenges-of-the-metaverse/
• https://shuftipro.com/blog/the-rising-concern-of-financial-crimes-in-the-metaverse-aml-screening-as-a-solution/

‍