From Mitnick to Modern Database Threats: Evolution of Social Engineering Tactics and Their Impact on Data Integrity

1. Introduction

The human aspect of computer security is frequently exploited and often neglected. Organizations invest millions in firewalls, encryption, and securing devices, but much of that effort is wasted if it doesn’t address the most vulnerable link in the security chain: the people who use, manage, and operate the systems (Mitnick & Simon 2003). With rapid technological advancement, the database and cyber security landscape has become increasingly complex and as primary repositories for sensitive personal, financial, and operational data, databases have become prime targets for cyber attackers (Li, Zhou & Cao 2021). Among the most dangerous tactics used is social engineering, which manipulates human behavior to bypass even the most advanced technical safeguards. Social engineering is particularly strong because it exploits the weakest link in cyber security which is human nature (Falade 2023). Kevin Mitnick, one of the most iconic figures in the history of social engineering, revolutionized how attackers approach database infiltration in the late 20th century. Unlike most hackers of his time, who focused on exploiting software or hardware vulnerabilities, Mitnick understood that humans could be more easily manipulated than code (Wang et al. 2020). His techniques as shown in Figure 1, such as pretexting and dumpster diving, enabled unauthorized access to major organizations like Motorola and Sun Microsystems. Mitnick’s legacy illustrates that, despite advancements in cyber security, humans remain a primary point of exploitation for adversaries, an insight that attackers continue to adapt today.

Figure 1

Kevin Mitnick’s early social engineering tactics, once reliant on simple human manipulation, have evolved into complex strategies that now target entire data ecosystems, including networks, cloud systems, and databases. Today, attackers use Artificial Intelligence (AI) and machine learning to automate and personalize social engineering techniques, such as phishing and pretexting, allowing them to impersonate trusted figures and extract sensitive information with ease (Chetioui et al. 2022; Kathiravan et al. 2023). Social engineering methods like baiting and quid pro quo have also advanced, exploiting human psychological triggers like curiosity and urgency. AI-driven algorithms enhance the effectiveness of these tactics, making modern attacks more efficient, harder to detect, and increasingly damaging, often leading to ransomware deployment, organizational disruption, and financial loss (Becue, Praça & Gama 2021). As database systems have grown to store vast amounts of sensitive information, they have become high-value targets for cybercriminals, with human operators, administrators, and users representing potential entry points for social engineers (Aldawood & Skinner 2020). Attackers now use sophisticated methods, such as spear-phishing, to target specific individuals with customized messages designed to extract credentials or gain unauthorized access (Kwak et al. 2020). This level of precision far surpasses the generalized tactics of Mitnick’s time. Modern attackers spend weeks or months researching their targets, creating believable narratives that significantly increase their chances of success.

Over the years, social engineering attackers have increasingly exploited human behavior to bypass advanced cybersecurity defenses. By leveraging psychological triggers like trust, authority, curiosity, and urgency, attackers have crafted increasingly sophisticated tactics to launch such attacks. As of 2023, social engineering attacks have reached a staggering 1.76 billion as shown in Figure 2 (Chheda 2024). One prominent example is the 2013 Target breach, where attackers phished a third-party vendor, leading to the theft of credit and debit cards from over 40 million customers, along with personal information from an additional 70 million customers. The breach resulted in an estimated financial loss of $292 million and Target faced severe reputational damage and legal challenges, including a $10 million class-action lawsuit settlement (Shu et al. 2017). In the 2016 Uber breach, attackers used vishing attack which is combination of voice and phishing to gain administrative credentials, exposing data of 57 million riders and drivers. The attackers accessed Uber’s internal systems, exfiltrating data on 57 million riders and drivers (Kiesow Cortez 2020). Uber initially concealed the breach and paid $100,000 to the attackers to destroy the data, which led to severe legal consequences and eventual settlements amounting to $148 million in fines for failing to disclose the incident (Coffey 2019). Similarly, in 2020, Twitter was compromised through phone-based phishing, where attackers hijacked high-profile accounts to promote a crypto currency scam, resulting in reputational damage (Kekulluoglu, Magdy & Vaniea 2020). These incidents underscore the growing danger of social engineering, which continues to exploit human vulnerabilities, causing significant financial and reputational damage to organizations globally.

Figure 2

This paper examines how social engineering tactics have evolved in response to modern global threats, emphasizing how attackers exploit human vulnerabilities to compromise data integrity within a database infrastructure. Despite technical and human safeguards, social engineering remains the most effective method for bypassing robust defenses by obtaining sensitive information like passwords or exploiting insider cooperation. The research traces the evolution of social engineering techniques and offers insights into the existing approach employed to counter such attacks. Furthermore, this paper also highlights how organizations can better defend against these sophisticated attacks. It stresses the need for cyber and database security awareness and training and use of access control and authentication to prevent social engineering attacks. Ultimately, the paper connects the historical context with emerging security strategies, encouraging organizations to adapt their defenses to counter these persistent threats.

2. Modern Social Engineering Tactics

As technology advances, so do the methods used by attackers to exploit human vulnerabilities. Modern social engineering tactics have evolved beyond basic deception due to integration of AI, machine learning (ML) and automation to bypass even the most robust security measures.

The integration of AI has now improved the way phishing and spear phishing tricked the victims. In past, it involved sending deceptive emails to trick individuals into revealing sensitive information. Now a days, with the use of AI, attackers employ context aware message using natural language processing and behavioral profiling to deceive the target and bypass spam filters (Jain & Gupta 2021). According to the FBI’s 2023 Internet Crime Report, phishing remains the most frequently reported cybercrime, with losses amounting to over $44 million (Chan-Tin & Stalans 2023). The 2016 DNC breach, where attackers used spear-phishing to steal thousands of emails, is a notorious example that demonstrates the potentially far-reaching consequences of these personalized attacks (Prevezianou 2021). There are different forms of social engineering attacks by researchers over the years as shown in Figure 3. Pretexting attacks are now more convincing than ever because of deep fake technologies which helps in the creation of a fabricated scenario (fake voice or fake video) designed to manipulate a target into revealing sensitive information or performing an action. Unlike phishing, which relies on creating a sense of urgency for quick decision-making, pretexting is more subtle, focusing on building trust over time (Mehta, Vora & Sachala 2021).

Figure 3

On the other hand, vishing relies on phone calls to impersonate legitimate entities, such as technical support or financial institutions, to extract confidential information. Now a days, the sophistication of these attacks are further improved by AI generated voices that can easy impersonate an organization executive or employees. Vishing attacks often result in the compromise of credentials, which can later be used to gain unauthorized access to sensitive databases or systems (Ashfaq et al. 2024). In baiting, attackers employ AI to integrate malware which is delivered through free software downloads or physical items like USB drives to bypass antivirus detection (Tulkarm 2021). One real-world example is from 2008, when attackers left infected USB drives in the parking lot of the US base. Curious employees plugged the devices into their work computers, unknowingly installing malware that allowed attackers to gain unauthorized access to sensitive defense systems. This breach led to increased scrutiny on physical security protocols across defense contractors (Roberts 2010).

Water holing attacks have not evolved into more targeted attacks with AI algorithm identifying which websites target groups should visit to inject a malicious code or malware (Ismail et al. 2017). In 2019, cybercriminals injected malicious code into legitimate websites frequented by individuals and organizations related to religious, charity, and volunteer sectors. Visitors to these compromised sites were unknowingly subjected to a drive-by download, where malware was silently installed on their devices without their knowledge (Paganini 2020). Furthermore, the quid pro quo attacks, where attackers offer services or benefits in exchange for sensitive information, are now carried out by AI based chat-bots posing as tech support in real time chat interfaces (Kamruzzaman et al. 2023). In 2018, an attacker posed as tech support for Arizona-based Banner Health, convincing employees to provide login credentials under the guise of system upgrades. The breach cost the company millions in compliance fines and forced significant operational changes (John, Ravichandran & Khan 2018).

Reverse social engineering involves deliberately creating a problem, after which the attacker poses as a solution provider to gain victim trust (Krombholz et al. 2015). Now a days, the attackers employ spoofed system errors and automated alerts generators to lure the victim into making contact. Even physical attacks like tailgating are harder to detect because they integrate behavioral profiling with physical deception to gain physical access to a restricted area.

These advancements indicate that social engineering has moved from traditional manual manipulation to more sophisticated, coordinated and data driven methods empowered by AI and ML. As attackers shift their methods towards automated behavioral profiling and exploit decision making biases, traditional methods for awareness and training to counter such attacks need to be supplemented with advanced behavior awareness systems and technical countermeasures.

3. Social Engineering Attacks Impact on Database Integrity

Social engineering attacks pose a significant threat to database integrity by exploiting the human element in organizational security frameworks. These attacks manipulate individuals into revealing sensitive information, such as login credentials to bypass authentication measures. Once attackers gain unauthorized access, they can manipulate, delete, or insert false data, compromising the accuracy and reliability of the database (Soni & Mathew 2020). This differs from traditional hacking, which exploits technical flaws, as social engineering undermines the trust users place in legitimate processes, making it harder to detect and prevent. For example, phishing attacks may lure employees into revealing passwords or downloading malicious software that allows attackers to alter database entries undetected. These manipulations can result in incorrect financial records, falsified personal information, or changes in access privileges leading to significant disruptions in data-driven decision-making (Jakala & Pekkola 2007). In the context of enterprise databases, social engineering attacks can have a cascading impact affecting not just the targeted system but also interconnected platforms and services reliant on accurate data. Once compromised, database integrity becomes a challenge to restore, often requiring detailed forensic analysis to uncover the extent of the manipulation. Organizations may face operational setbacks, loss of reputation, legal penalties, and diminished customer trust, especially if the attack goes undetected for an extended period (Neto et al. 2021). In this section, we explore different case studies of known social engineering attacks, their consequences and also explore the consequences of database integrity compromise.

3.1 Case studies of notable breaches and their consequences

• Sony Pictures Hack (2014): Attackers used phishing to lure Sony employees into revealing their credentials, which allowed the attackers to infiltrate the corporate network. They then exfiltrated sensitive emails, internal documents, and unreleased films while erasing key data from Sony’s servers. Although the primary objective was not to alter the integrity of Sony’s databases, the incident underscored the broader risks social engineering presents to data integrity. Figure 4 indicates the impact of Sony pictures breach on its revenue (Haggard & Lindsay 2015; Sicard 2015; Quader & Janeja 2021).

• The Snowden Revelations (2013): Snowden, a contractor for the National Security Agency (NSA), used social engineering to persuade colleagues to reveal their login credentials. Armed with these credentials, he accessed and exfiltrated vast amounts of classified data (Boussios 2021). The case revealed how trusted insiders, leveraging social engineering, could become significant threats to database integrity.

• Toyota Boshoku: Attackers defrauded Toyota Boshoku Corporation of $37 million through a business email compromise (BEC) scam by persuading a finance executive to redirect wire transfers to falsified account details (Gatefy). The case demonstrates how attackers can manipulate data, including database records tied to financial transactions.

• OCBC Bank Phishing Scam (2021): This case involved the Oversea-Chinese Banking Corporation (OCBC), where phishing attacks led to the loss of $8.5 million from 470 customers. Attackers created fraudulent websites mimicking the bank’s interface, tricking users into revealing login credentials. Figure 5 highlights the impact of OCBC Bank Phishing Scam.

• Colonial Pipeline Attack: A ransomware attack enabled through social engineering tactics resulted in shutdown of fuel supply across the East Coast. The company reportedly paid approximately $4.4 million in ransom to regain control of its systems (Beerman et al. 2023; Pitman & Crosier 2024).

Figure 4

Figure 5

These cases highlight how social engineering attacks, particularly those targeting human errors, can lead to data compromise in a database system. By gaining unauthorized access, attackers can alter or delete key records, impacting data accuracy and trust. Companies facing such attacks must deal not only with the immediate financial and reputational damage but also with long-term challenges in restoring database integrity.

3.2 Consequences of database integrity compromise

The consequences of social engineering attacks on database integrity are severe and multifaceted, impacting organizations across various dimensions, including operational efficiency, legal liabilities, financial stability, and overall reputation as depicted in Figure 6. Importantly, these breaches also raise significant concerns related to national security, particularly when sensitive governmental or critical infrastructure data is compromised.

Figure 6

3.2.1 Operational disruption

When data is corrupted or manipulated, organizations struggle to maintain normal business functions. In critical sectors such as healthcare, an erroneous database could lead to misdiagnoses or improper patient care. For national security, disruptions in critical infrastructure such as energy, transportation, and emergency services can have devastating effects on public safety and security (Orlando 2021; Patrick, Brett van Niekerk & Fields 2019).

3.2.2 Legal and financial repercussions

Data integrity breaches expose organizations to significant legal challenges, and they may face lawsuits from affected individuals or businesses, especially if personal information is altered, compromised, or lost. For example, under GDPR, organizations can face fines of up to €20 million or 4% of their global annual turnover for failing to protect personal data (Başak Erdogan 2021; Hoofnagle, Sloot & Borgesius 2019; Presthus & Sonslien 2021).

3.2.3 Reputation damage

The fallout from a data breach often includes severe damage to an organization’s reputation. Once a breach is publicized, customer trust can erode quickly, as stakeholders question the organization’s commitment to safeguarding sensitive information. A good example is the 2017 Equifax breach, which exposed the personal data of approximately 147 million individuals, resulting in a significant reputational harm, with many customers choosing to sever ties with the company (Perera et al. 2022; Novak & Vilceanu 2019).

3.2.4 Difficulty in restoring data

The process of restoring database integrity following a breach is typically complex and resource-intensive. Organizations often engage forensic experts to conduct thorough investigations, assess the extent of the damage, and determine the vulnerabilities that allowed the breach to occur (Bansal 2019). In sectors related to national security, such as intelligence or defense, even minor inaccuracies in data can lead to grave consequences, potentially jeopardizing national interests or security operations. The time taken to restore integrity can lead to prolonged operational setbacks, impacting response times in critical situations (Hathaway & Klimburg 2012).

3.2.5 Erosion of competitive advantage

A compromised database can diminish an organization’s competitive advantage in the marketplace, especially for companies engaged in national security or defense contracts. Businesses known for their commitment to data security may find their reputations tarnished by breaches, leading competitors to capitalize on their vulnerabilities. The loss of contracts with governmental bodies can have cascading effects, affecting not just the organization but also the national supply chain for critical goods and services (Morse, Raval & Wingender 2011).

3.2.6 Increased security costs

Following a data breach, organizations often face heightened security costs as they scramble to bolster their defenses against future attacks. This may involve hiring additional data security experts, investing in advanced security technologies, and conducting regular security audits. For organizations linked to national security, the stakes are even higher, as failures in cybersecurity can lead to threats against public safety (Algarni & Malaiya 2016).

3.2.7 National security implications

When social engineering attacks target organizations that handle sensitive governmental or critical database infrastructure, the implications extend to national security. Compromised databases may contain data intelligence, defense strategies, or infrastructure plans vital to national interests. For instance, a successful breach at a defense contractor could provide adversaries with insight into military operations, potentially undermining national defense (Smith & Mulrain 2017).

4. Methodology

This section outlines the research methods and systematic approach used to conduct the work on the evolution of social engineering tactics and their impact. The goal is to distinguish the patterns and assess how modern technologies such as artificial intelligence (AI) and machine learning (ML) have improved the efficiency of social engineering attacks.

5. Existing Technique to Mitigate Social Engineering Attacks

Social engineering attacks exploit human psychology, making them one of the most challenging forms of cyber threats to defend against. As these attacks increasingly target organizations’ most vulnerable asset people various strategies have been developed to mitigate their impact. Butt et al. (2022) proposed a system to detect phishing emails using machine learning and deep learning models, specifically Support Vector Machine (VM), Naive Bayes (NB), and LSTM. They created a dataset with legitimate and phishing emails, applied feature extraction techniques, and trained the models for classification. Their approach achieved high accuracy rates, with SVM reaching 99.62%, NB 97%, and LSTM 98%. They suggested improving the system by combining phishing and legitimate emails into a more dynamic dataset to better reflect real-world conditions. The proposed framework aims to enhance protection against evolving phishing attacks across organizations. Another technique to detect phishing attack was proposed by (Salahdine, El Mrabet & Kaabouch 2021), in which they utilized machine learning-based technique for detecting phishing attacks, focusing on emails. They collected and analyzed over 4000 phishing emails and built a dataset with 10 relevant features. The dataset was used to train, validate, and test three classifiers: Support Vector Machine (SVM), Logistic Regression (LR), and Artificial Neural Network (ANN). Through parametric studies, they found that SVM, LR, and ANN achieved high accuracy, with the ANN model performing best using two hidden layers and the Relu activation function. This model offers fast and accurate phishing detection. Benavides-Astudillo et al. (2023) proposed a phishing detection model that focuses on analyzing the text of suspicious web pages, rather than just URLs, using Natural Language Processing (NLP), and Deep Learning (DL) algorithms. They applied the Keras Embedding Layer with GloVe to capture semantic and syntactic features of web page content and tested four DL algorithms: LSTM, BiLSTM, GRU, and BiGRU. Their experiments showed that BiGRU performed the best, achieving a mean accuracy of 97.39%. The study highlights the potential of using word embeddings and DL algorithms for effective phishing detection.

To deal with sub type of phishing attack called vishing, Fakieh & Akremi (2022) proposed blockchain technology-based mechanism to enhance security. Their model is based on two key principles: dual control (requiring more than one entity to complete a task) and split knowledge (dividing critical information between entities). By incorporating blockchain, the model ensures secure, transparent, and immutable records of transactions, reducing reliance on human vulnerability. The paper highlights its effectiveness in handling vishing attacks and demonstrates how the model can be easily adopted by organizations using open-source blockchain platforms. Al-Hamar et al. (2021) put forward a solution to detect spear-phishing attacks by analyzing domain authenticity, which attackers often manipulate to gain trust. Their detection system uses multiple algorithms to evaluate whether a domain is genuine or fraudulent, significantly improving phishing detection. Tests conducted showed it outperformed existing email security systems, in identifying spear-phishing and whaling attacks. Through real-world testing, their work successfully detected phishing emails that other systems failed to recognize, demonstrating its potential to enhance protection against enterprise-targeted attacks. Sonowal et al. (2021) took an approach to handle traditional phishing and spear-phishing emails using multi-dimensional features such as domain validation, writing style, and an auto-upgrade profile is proposed. The method employs a machine-learning algorithm to classify phishing emails and includes an additional layer of security through a verifiable secret-sharing scheme to authenticate email senders. This approach effectively enhances the detection of phishing attacks by leveraging various features and adding a unique verification step.

Researchers have also worked over the years to handle watering hole attacks. Allen et al. (2020) proposes Mnemosyne, a forensic analysis engine designed to investigate watering hole attacks, where attackers compromise a frequently visited website to infiltrate an organization’s network. Mnemosyne passively collects browser activity logs to detect when and how a website was compromised. It then traces the impact of the attack on specific users within the targeted enterprise. Through real-world tests, the system effectively identified victims and reduced manual analysis effort by over 98%. Mnemosyne enhances post-attack investigations, pinpointing the attack timeline and affected employees. Irani et al. (2011) examine reverse social engineering (RSE) attacks within social networks, where the victim initiates contact with the attacker, establishing trust. Unlike traditional social engineering, RSE manipulates features like friend recommendations, visitor tracking, and demographic-based searches to lure users. Their work discusses three novel RSE attack strategies, demonstrating their effectiveness in real-world scenarios. Their work also shows that attackers can attract victims without sending friend requests, simply by creating enticing profiles. The study highlights the need for social networks to implement countermeasures against RSE threats.

Tailgating attack, which is a form of social engineering attack, poses significant threat to physical security. To handle this, Akati & Conrad (2021) proposed a three-step anti-tailgating solution to address the shortcomings of existing security technologies. This system combines face detection, palm recognition, and motion sensors to detect multiple individuals or fast-paced entries that bypass single access authorizations. The design science methodology and experimental approaches were used to develop and test this solution. Results demonstrated that the combination of these technologies effectively prevents tailgating by overcoming limitations of prior methods. This system significantly enhances security for organizations. Girinoto et al. (2022) explored how phishing attacks could be executed by pretexting within the OmeTV video chat application, targeting users from Indonesia over 18. They proposed a Social Engineering Session (SES) attack methodology to extract personal information from victims by building trust. The results showed that the attack successfully gathered sensitive details, such as full name, date of birth, address, hobbies, Instagram account, and phone number. They also suggested that this information could be further exploited for fraud or marketing purposes. The study highlights the effectiveness of pretexting in phishing attacks. A detailed comparison of state-of-the-art method used for social engineering attacks are shown in Table 1.

6. Measures to Protect from Social Engineering Attacks

Organizations must educate and provide training to their employees on recognizing and responding to social engineering tactics, as human awareness remains the first line of defense. Regular cyber security awareness programs can help staff identify phishing, pretexting, baiting, and other deceptive schemes that exploit human behavior. In addition to education, organizations should implement multi-layered security protocols, such as two-factor authentication and strict access controls, to reduce the chances of successful attacks. Continuous monitoring, alongside incident response simulations, can also prepare employees to act swiftly in case of a breach. A comprehensive approach that combines education, technological safeguards, and policy enforcement is critical to mitigating the growing threat of social engineering attacks.

In their work, Weaver et al. (2021) emphasized that effective training should guide users on specific elements to look for when determining if an email is phishing. These cues include scrutinizing the sender’s email address to ensure it matches the legitimate domain of the organization, checking for subtle discrepancies in the URL of links by hovering over them to preview the actual destination, and watching for common red flags like urgent or threatening language aimed at creating a sense of panic. Users were also trained to identify unusual requests for personal or financial information, especially in unsolicited emails. Additionally, they were encouraged to look for inconsistencies in the email format, spelling or grammar errors, and unfamiliar logos or poor-quality images. By teaching users how to verify these indicators, the training helped them more confidently classify emails as phishing or legitimate, even when phishing emails appeared highly sophisticated. The study conducted by Thomas (2018) emphasized the importance of personalized training tailored to different user groups to improve phishing detection and enhance overall security awareness within organizations. The author proposed a method to identify phishing attacks by conducting interviews with seven security experts. The authors conducted interviews with seven security experts to explore methods for empowering users to resist phishing attacks. They identified nine key themes that help distinguish users who are vulnerable to phishing from those who are more resilient. These themes include awareness of cybersecurity threats, email analysis skills, caution with unsolicited communications, trust in organizational policies, personal experience with phishing, technical proficiency, confidence in judgment, frequency of training, and responsiveness to authority. By addressing these factors, organizations can enhance phishing detection and improve overall cybersecurity awareness, reducing the likelihood of successful attacks.

According to Salahdine & Kaabouch (2019), to prevent tailgating attacks, awareness programs and training should be introduced on the importance of not allowing unauthorized individuals to follow them into secure areas, even out of politeness. Training should emphasize strict adherence to access protocols, such as ensuring each person uses their own credentials and reporting suspicious behavior. Employees can also be taught to challenge unknown individuals or direct them to security personnel. Regular awareness campaigns and physical security drills can reinforce this behavior. Training should highlight real-life examples of tailgating incidents and their consequences.

Vadrevu & Perdisci (2019) discusses that flow whitelisting is a proactive security measure that can effectively prevent social engineering attacks if employed by organizations. By restricting the actions and access of users within a system and by defining a set of trusted actions and entities, it limits unauthorized interactions that attackers might exploit. This approach reduces the attack surface, making it more difficult for social engineers to execute their attack. Additionally, monitoring user behavior against established patterns can help identify suspicious activities, triggering alerts for potential threats.

According to Montanez et al. (2022), access control is crucial in minimizing social engineering attacks by limiting who can access sensitive information and systems. By implementing role-based access controls (RBAC), organizations can ensure that employees only have access to the information necessary for their specific job functions. This reduces the chances of an attacker obtaining sensitive data through manipulation or deception. Additionally, Siadati et al. (2017) states enforcing multi-factor authentication adds another layer of security, making it more difficult for unauthorized individuals to gain access, even if they manage to deceive an employee into revealing their credentials. Overall, robust access control measures help create barriers that protect against exploitation by social engineers. Table 2 provides a deep insight into measures employed against social engineering attacks.

7. Future Directions

As social engineering attacks are continuously evolving, social engineering attackers are exploiting human psychology, system vulnerabilities, and advanced technologies to launch more sophisticated and targeted attacks. It is necessary that future work should focus on the following directions to handle the increasing threats to database integrity.

• In the future, researchers should employ machine learning methods for the detection and mitigation of anomalies that indicate social engineering attacks. Natural language processing (NLP) can also be incorporated into voice recognition systems and email filters to detect suspicious activities, in particular phishing and vishing attempts.

• Conventional access control systems fail against credential-based attacks. Future work should use decentralized identity systems and continuous authentication methods that employ biometrics and behavioral patterns. This will help minimize the damage even when credentials are compromised through social engineering.

• With the advancement in AI, attackers generate fake media in pretexting and impersonation. To overcome that, researchers should design plug-and-play verification tools that are able to integrate with organizations’ video conferencing, email, and HR systems to detect manipulated content.

• The human remains the most vulnerable element. Future research should integrate cognitive behavioral insights into training programs, transforming awareness into behavioral resistance.

• Social engineering attacks also take advantage of interconnected systems across different sectors. In the future, policy and infrastructure should provide robust, anonymized sharing of threat data between organizations to improve the defense against evolving social emerging methods.

• In the majority of cases, the breaches remain poorly documented or underreported. Future researchers should focus on the development of standardized breach taxonomies that not only classify the technical vectors but also the components of social engineering. Such work will help improve public awareness and large-scale analysis.

8. Conclusion

Kevin Mitnick’s pioneering social engineering tactics revealed the profound vulnerability of human nature in cyber security. Since then, attackers have significantly advanced their strategies, blending psychological manipulation with sophisticated technologies to target database integrity with increasing precision. In this study, we have explored in depth various types of social engineering attacks, illustrating their impact through real-world examples. Furthermore, we examined existing techniques, including machine learning-based detection systems and block chain, for securing access alongside preventive measures that individuals and organizations should adopt to mitigate the risks posed by social engineers. We also highlighted the limitations of these existing approaches and explored potential future directions that should be employed to handle the ever-evolving threat of social engineering attacks. Ultimately, this study highlights the critical need for comprehensive and adaptive defenses, combining technology, human awareness, and continuous innovation to safeguard databases and prevent future breaches.

Data Accessibility Statement

The data analysed in this study is taken from publicly available academic sources, industry reports, white papers, and documented cases of incidents and breaches. A full list of sources is included in the references. Detailed indexing criteria and source material are available upon request.

Competing Interests

The author has no competing interests to declare.