Artifical Intelligence in Cryptography: Protection of Cryptographic System under Patents

Paper Code: AIJACLAV10RP2025

Category: Research Paper

Date of Publication: May 19, 2025

Citation: Ms. Kummasani Vinodhini, “Artifical Intelligence in Cryptography: Protection of Cryptographic System under Patents", 5, AIJACLA, 95, 95-111 (2025), <https://www.aequivic.in/post/artifical-intelligence-in-cryptography-protection-of-cryptographic-system-under-patents>

Author Details: Ms. Kummasani Vinodhini, LLM Scholar, GITAM (deemed to be) University

Download PDF

Abstract

Artificial Intelligence (AI) is transforming cryptographic approaches to encryption techniques, automating key generation, and strengthening safety mechanisms against cyber threats. Possible applications include strengthening data protection, detecting vulnerabilities, and optimizing cryptographic protocols. In particular, AI-based cryptographic models enhance the non- disable for security systems against advanced assaults, including those posed by quantum computing. Cryptography leverages AI through machine learning algorithms for anomaly detection, AI-based key management, and neural network-based encryption techniques. Basically, these advancements allow for adaptive security frameworks that can recognize and take action against potential breaches in real time. AI-assisted cryptanalysis can also be used to assess the strength of encryption, therefore restructuring cryptographic practices. Legal protection for AI-driven cryptographic solutions under intellectual property laws, particularly the Patents Act, will remain a crucial area of concern as they continue to evolve. Patentability in cryptography has faced challenges because it relies on mathematical algorithms, which are often not considered patentable in many jurisdictions. However, when technical applications, such as higher efficiency or enhanced security, can be demonstrated by AI-based cryptographic methods, they may find a place under the patent-protectable framework. Novelty, inventive step, and industrial applicability are three of the criteria AI innovations in cryptography must satisfy under the laws of patents, which include the Indian Patents Act, 1970, and European Patent Convention (EPC). The patent framework for cryptography is now developing, with courts and regulatory authorities analyzing the degree of protection that should be accorded to cryptographic mechanisms arising from AI. The paper analyses how AI and cryptography intersect regarding patentability in AI-powered cryptographic systems and the challenges posed by the existing intellectual property frameworks.

INTRODUCTION

Cryptography involves both the study and practice of techniques for secure communication in such a way that third parties, called adversaries, can neither detect nor maliciously decrypt or interpret, or pose a threat to the confidentiality of messages shared by two parties. It turns to develop and analyze protocols that should ensure that attackers can’t retrieve information being shared between two entities, thus ensuring various aspects of information security. Secure communication is the situation where the message or data shared between two parties can’t be viewed by an adversary. In Cryptography, this Adversary is a malicious one that aims to retrieve precious information or data there by violating those aspects of information security. Data confidentiality, data integrity, authentication, and non-repudiation are the fundamental principles of present-day.

●        Confidentiality generally means some rules that have guidelines and conventions under confidentiality agreements that fix the usage of information to specific people or place.

●        Data integrity is the maintenance of the whole life cycle of data in an accurate and consistent manner.

●        Authentication is to make certain that the data being asserted by the user belongs to them.

●        Non-repudiation is the ability to ensure that a person or party to a contract, or communication, is unable to deny having signed the paper or sent the message.[1]

HISTORY

The practice of keeping information secret while at the same time communicating it to a small number of people through the manufacture of codes and ciphers has gone through major developments in a millennium to influence technological advancements and legal frameworks.

Medieval Cryptography (1000-1500 AD)

The art of cryptography dates back to ancient civilization, but its popular use-as we know it-began during the 1000 CE period. Among the Arabs during this period, the development of complex codes for concealment of communication from potentially malicious actors was recorded for the first time. Of the many, the most notable was the so-called Caesar cipher, a class of substitution cipher first used by Julius Caesar to encrypt messages by shifting the letters of a phrase some prescribed space along the alphabet.

In the ninth century, Al-Kindi, an Arab mathematician, wrote the book in which he introduced techniques for breaking simple ciphers through frequency analysis. This activity marked the advent of cryptanalysis, laying the grounds for future advances in cryptography.[2]

Renaissance and Early Modern Cryptography (1500-1800 AD)

As the Renaissance commenced in Europe, cryptography gained a more complex foothold. From simple codes, it passed into the hands of Leon Battista Alberti, an Italian Renaissance polymath inventor of the Polybius square and, later, the cipher disk, a system of encryption using substitution ciphers. Thus, a transition occurred from simple ciphers to more compact systems.

In the sixteenth century, the Vigenère cipher was formulated by Giovan Battista Della Porta and Blaise de Vigenère. This cipher improved coding through a keyword for encryption but found further development in the capacity for breakage of the code. However, these techniques were still exposed to vulnerability by classical cryptanalysis.[3]

19th Century: Industrial Revolution and Advances in Cryptography

In the 19th century, mechanical encryption devices began to emerge. The Jefferson disk cipher first created by Thomas Jefferson in 1795 was an early effort to automate encrypting messages through a series of rotating disks’

With wars came the centrality of cryptography, particularly with the emergence of machines.

A prevalent example was the Enigma machine famously used by Nazi Germany and employing a complex arrangement of rotors defused to constitute a very strong code. Nonetheless, the success of British cryptanalysts in breaking the Enigma code, particularly through the efforts of Alan Turing, marked a turning point in the history of cryptography.[4]

Modern Cryptography (20thCentury-Present)

The 20th century witnessed mammoth changes in cryptography following the invention of computers. Thus, started in the 1970s the beginning of modern symmetric encryption with the acceptance of the Data Encryption Standard. The very development of Public Key Cryptography in 1976 by Whitfield Diffie and Martin Hellman followed this, which established algorithms called RSA after its originators Rivest, Shamir, and Adleman. Public-key cryptography brought the concept of two keys, one public and one private to be shared among people so that they can communicate securely, thus resolving the issue of key distribution in earlier cryptography systems.

Patent Protection for Cryptographic Systems

The protection of cryptographic systems by patents has been very influential in the way modern encryption technology has grown. In the mid-1970s through 1980s, patents were central to the evolution of encryption algorithms, and many early encryption techniques, such as DES and RSA, were patented. Patents served to secure intellectual property rights for cryptographic systems, providing motivation toward innovation. For instance, in 1977, RSA was patented by its inventors and the patent provided them control over its commercial use until its expiration in 2000.

However, the patenting of a cryptographic algorithm also raises concerns against its security. Patented algorithms are kept usually confidential and therefore pose an interference towards independent analysis of their strength and weaknesses. On the other hand, open-source cryptographic algorithms, e.g., those based on the AES (Advanced Encryption Standard), have been widely used in modern systems and are open to the public for scrutiny.

Current position

Today, cryptography occupies the core of securing digital communication. It ranges from end-to-end encryption in messaging apps, to blockchain, to cryptocurrency technologies. The advent of quantum computing presents new challenges, as classical cryptographic systems could be broken by quantum algorithms, raising the need for the emergence of post-quantum cryptography that is, the new area of research which focuses on developing new cryptographic methods not able to be broken by quantum-based attacks.[5]

Though patents are contentious in cryptography, the arguments dissect the issue of how they affect accessibility and innovativeness. The open-source community ensures that its algorithms remain unpatented and transparent, thereby avoiding a clash between security and passing off through intellectual property rights.

In a nutshell, this thousand-year development of cryptography into various expressions is not only technological but also largely due to the influence of patent regulation. Although being a core player in the securing of communication and data, the prospects for cryptography would mainly depend on how ready it is to take new challenges from quantum computing and how its position would evolve on striking a balance between intellectual property protection and open security standards.

CRYPTOGRAPHIC SYSTEM

A cryptosystem is a composite of a set of algorithms that encode or decode messages so they can be readable only to authorized users. The term cryptosystem is shortened from "cryptographic system" which combines planning or building computer systems with cryptography that is protecting information and communication by codes so that only veritable receivers for whom, information is on offer, can realize and process it.

These algorithms provide the capability for key generation, encryption, and decryption. The cryptographic key is a string of bits used by the cryptographic algorithm to carry out the transformation of plaintext to ciphertext and vice versa. It is a part of the variable data fed into a cryptographic algorithm to perform such an operation. The security of the cryptographic scheme highly depends on the protection of the keys employed.

The cryptosystem is used for the transmission of messages hypothesised by bank cards, credit cards, and private data over the internet for one such possible application. In a different area under the heading of cryptography, secure email would help embody methods for the establishment of digital signatures, cryptographic hash functions, and key management techniques.[6]

ROLE OF AI IN CRYPTOGRAPHY

Integrating AI in cryptography transforms cyber security and stands as strength permitting the encoding of encrypted messages. Many functions of AI exist such as diminishing the generation and management of encryption in algorithms all immune to attacks. The activity of securing communication and information through a system of codes is the role of cryptography. AI means a lot, and with machine learning, deep learning, and other techniques, and stands a chance of being able to analyze in which there are scopes for strengthening security measures in cryptographic systems. AI for strong cryptographic techniques can automate processes such as key generation, protocol verification, intrusion detection, and others, and thus offer strength and efficiency. Among these are some challenges faced by AI, includes adversaries breaking cryptographic codes-some new risks for digital security.

AI provides a novel tool to seal the digital data from threats. Therefore, collaboration of this field includes mathematics, cybersecurity, and law-not all topics are relevant to technical concepts.[7]

PATENT PROTECTION IN CRYPTOGRAPHY

The patentability area that involves the cryptography is allergy complicated. It would seem that, amongst other things, the advertising of cryptography-related patents is extremely complicated because most cryptographic algorithms-invariably implemented in software-are inherently mathematical and normally fall outside patentability criteria in many jurisdictions. However, ingenious applications or implementations of cryptographic techniques may still be patented, provided that they should give an adequate technical solution to a problem and not any inherently mathematical or algorithmic[8].

This paper considers the ways of AI-driven innovations in cryptography that cause ripples in encryption techniques and the problems of getting these innovations protected under patent laws of patentability, IP rights, and the changing legal landscape for safeguarding cryptographic technologies based on AI.

AI in CRYPTOGRAPHIC ANALYSIS

The branch of AI that has been said to be efficient for cryptanalysis is machine learning. Most of the techniques used to break the classical ciphers were performed either by hand or through brute force. Machine learning algorithms can do it much more quickly and much more effectively, looking for correlations and peculiarities of encrypted messages that human analysts could hardly find. In this way the neural networks can be trained on large databases with representatives of the employed plaintexts and their respective ciphertexts to generalize the pattern captured in turn to help decrypt any newly-generated cyphertext. Additionally, this feature improves the chances of successful decryption of classical ciphers and accelerates cryptanalysis[9].

AI in Side-Channel Attacks

In this context, side-channel attacks use any physical signal arising from the cryptographic device: time, power consumed or electromagnetic signals, to study the secret key. Besides, AI can multiply several times the efficiency of such an assault. However, in contrast to conventional means, machine learning algorithms are specifically able to analyze the side-channel data and reveal tiny correlation and pattern, which in return indicate the presence of cryptographic keys. Thus, on the other side of the coin, using AI side-channel attacks can be done in better and efficient way and it is really a hole in the cryptographic system.[10]

AI- Optimized Quantum Key Distribution

Quantum Key Distribution (QKD) is a method for establishing a secure key that works based on quantum mechanical principles. However, even though people consider QKD to have a pragmatic level of security that cannot be compromised, practical problems prevent its use: for instance, key management, together with error correction. AI can better control and enhance several of the aspects of QKD protocol set. Consequently, the keys generated through it are safe because AI has helped continuously correct for errors arising from quantum noise and other interferences. In that regard, this switch gives a general elevation in robustness and reliability for QKD systems.[11]

AI in CRYPTOGRAPHIC ALGORITHMS DESIGNS

An algorithm is defined as a set of instructions to be followed in the calculations or operations involving some mathematics and computer science. At the basic level, an AI algorithm is programs that tell the computer how to learn to work on its own.

An AI algorithm is much more complex than what most people are taught in algebra. Basically, AI programs are driven by a complex set of rules that define how they develop and learn. There would be no AI without an algorithm[12].

AI is at the forefront of cryptographic algorithm design by analysing huge amounts of data to identify patterns and weaknesses, which gives rise to robust and efficient encryption techniques, especially in areas like adaptive cryptography, key generation, and preparation against post-quantum computing attacks, where AI can assist in identifying the hardest algorithms susceptible to attacks. There are several keys characterises used in by AI in cryptographic algorithm designs.

A.      Identifying vulnerabilities: 

AI can analyze existing cryptographic algorithms to identify potential weaknesses or patterns that attackers can exploit. These weaknesses are improved in existing mechanisms.

B.      Adaptive cryptography:

AI can enable monitoring of data and network conditions in real time, so that adjustments can be made thereto on the basis of the risk perceived and the threat level posed. This gives an asymmetrical advantage, in security terms, in an adaptable way.

C.       Optimized key generation:

AI can assist in generating strong cryptographic keys through analysis that allows for more randomness in the process of user pattern identification, making brute-force attacks harder to run.

D.      Development of post-quantum cryptography:

This scenario is that one day quantum might be a threat. AI can be used to evaluate and come up with new encryption algorithms that can withstand any attack from quantum computers.

E.       Machine learning-based encryption:

AI can train new encryption algorithms that will be more complex and difficult to break on vast datasets using the learning models of machine learning.

F.       Security analysis and threat detection:

 By analyzing network traffic and suspicious patterns that hint at an attack, AI systems can allow an early detection of threats[13].

AI ENCHANCING CRYPTOGRAPHIC SYSTEM PROTECTION

Artificial intelligence is taking up a more vital role as far as improving the protection of cryptographic systems is concerned. Traditional cryptography relies on complex mathematical algorithms to secure information, with the progression of cyber-attacks making the demand for a dynamic and adaptive security system more pertinent. AI technologies, especially machine learning, deep learning, and neural networks, offer smartness, thereby enhancing cryptographic systems' strength, efficiency, and resistance against fill various attacks[14]. They are the major areas where AI has made a significant contribution to the improvements in the field of cryptography:

1. Automated Cryptanalysis and Vulnerability Detection

Cryptanalysis is the process of breaking or analyzing cryptographic systems. AI speeds up cryptanalysis by automating the detection of vulnerabilities in cryptographic algorithms, which, in a traditional way, can take a lot of time and effort. Whereas these methods may be slow and inefficient, algorithms such as AI, particularly machine learning models, learn to easily identify patterns or weaknesses in cryptographic algorithms that may be exploited by attackers. Automation will hasten the discovery and patching of vulnerabilities before attack by focusing on attack patterns.

For example, AI models could learn from huge datasets of previously unearthed weaknesses in cryptography, and that predictive capability could be harnessed to reveal potential new attack vectors, the results being more robust cryptographic systems.

2. AI-Facilitated Key Management and Protection

Key management is central to cryptography security; it involves the generation, distribution, storage, and management of cryptographic keys used in message encryption and decryption. When key management fails, the whole system's security is in jeopardy. AI facilitates key management in its generation, identification of unauthorized access, and detection of irregular usage patterns.

AI-Hypertext model can analyze the entire patterns of access to keys and can immediately elevate any suspicious behaviour. They are trained to win intelligence in analyzing such patterns and will improve their ability to detect passive keys over time.

3. Post-Quantum Cryptography

The advent of quantum computing poses serious threats to traditional cryptography, especially RSA and ECC, which a sufficiently powerful quantum computer would exploit. The power of quantum computing is propelling the race for PQC algorithms that would be robust to attacks based on quantum computing.

AI is being used to help design and optimize post-quantum cryptographic algorithms. Machine learning helps identify possible weaknesses in all PQC algorithms to predict which may hold up in a post-quantum environment. AI is likely to speed up the entire search and development of quantum attack-resilient algorithms to keep the cryptographic system secure in a world where quantum computers pose any danger

4. Intrusion Detection and Prevention

AI enhances the ability to detect and warn against criminal access in cryptographic systems through intelligent intrusion detection systems (IDS). Using the techniques of machine learning, these systems monitor network traffic for abnormal patterns or behaviors that may serve as indicators of an attack. An AI-driven IDS can even detect new attacks whose onset cannot be predicted based solely on the historical records by examining vast data networks or attack datasets.

Real-time analysis by deep-learning-based models becomes the perfect way of achieving quicker and more precise detection of breach or vulnerability of the cryptographic system. Such early detection aids the intended authority to block possible intrusion before such violators get into the cryptographic system.

5. Ensuring Blockchain and Cryptocurrency Security

AI performs an important function in securing blockchain and cryptocurrency systems based on cryptographic techniques that are sometimes complex. This improves the security of their protocols by optimizing mutual agreement mechanisms and detecting fraud or malicious activity, such as double spending and Sybil attacks.

AI can be construed as technology used for monitoring unusual activities on the blockchain network, predicting threats, and taking preventive measures for the cryptographic security of

information. For example, AI algorithms can analyze transaction habits, identify irregularities, and counteract them in an effort to minimize attacks on cryptocurrency systems.

6. AI for Optimizing Cryptographic Protocols

AI may optimize cryptographic protocols-such as secure key exchange and encryption algorithms-by adjusting parameters to strike the optimal balance between security and performance. AI can automatically adapt cryptographic protocols based on the gravity of the data, increasing the level of security without a dramatic decline in efficiency.

The use of machine-learning algorithms to perform analytics is an effective route to refining cryptographic schemes, so one such scheme could well adjust the encryption level according to real-time risks and usage.

AI in Securing Digital Communication

AI technology in cryptography usually works by maximizing the security of digital communication systems through extensive use of machine learning algorithms to discern patterns, find possible vulnerabilities, and develop stronger encryption techniques, thus forming a stronger cryptographic protocol that is less prone to attacks than conventional techniques; AI essentially becomes an advanced tool for both protecting and analyzing encrypted data, thus improving the overall strength of digital communication.[15]

1.       Smart Threat Spotting:

AI has the ability to look at huge amounts of info to spot tricky patterns and odd things that might show bad stuff going on. This lets it catch sneaky cyber-attacks that normal ways might not see.

2.       Changing Encryption:

By learning from info as it comes in, AI can change how it encrypts stuff on the fly. This helps it deal with new threats and makes the whole system tougher to crack.

3.       Watching How People Act:

AI keeps an eye on what users do and what's happening on the network. It can spot weird patterns that could mean someone's trying to break in. This lets people fight back before things get bad.

4.       Private Machine Learning:

AI can do machine learning on secret info while keeping users' privacy safe. It uses tricks like making the data fuzzy or doing math on encrypted stuff to pull this off.

PATENT PROTECTION FOR AI-DRIVEN CRYPTOGRPAHIC SYSTEMS

To secure a patent for AI-powered encryption tech, you got to spotlight how the AI's mix into the code-cracking methods stands out and isn’t just something anyone could think up. Point out the cool tech upgrades you're bringing to the table and lay out the formulas and uses baked into the setup. Going this route makes sure your brainchild ticks the boxes for being fresh, not a no-brainer, and useful—it also tackles the tough bits of getting legal rights for inventions that are all about that software life[16].

●        Focus on the application, not just the AI algorithm: Even though you might not get a patent for the AI technique itself, the method you use it to crack a certain crypto puzzle in the system could be under legal protection.

●        Technical contribution: You got to show how the AI steps up the game for the crypto system maybe by making it safer, speedier, or better at handling new dangers.

●        Claiming the inventive aspects: Dig out what sets your AI-powered crypto system apart, like the exact type of machine learning it uses, the data you train it on, and how it makes choices.

●        Addressing legal challenges: Remember patent offices might give you a hard time with your AI-related claims because they can run into problems like being too "theoretical" or just "number stuff."[17].

Patenting AI Cryptography Innovations

Artificial Intelligence (AI) and cryptography are two of the most transformative technologies of the 21st century. When combined, they can lead to groundbreaking innovations that promise to revolutionize fields ranging from cybersecurity to blockchain. However, as AI continues to evolve and drive change within cryptography, the question of patenting these innovations becomes increasingly complex. This article explores the patenting of AI-based cryptographic innovations, the challenges associated with patent protection, and the legal precedents and global patent laws that impact the process.

Understanding AI Cryptography Innovations

AI and cryptography smashed together mean using smart computer stuff to make secret code systems better and safer. They're doing stuff like:

1. AI-Boosted Secret Codes: They're teaching machines to make secret codes that toughen up and change when baddies get clever.

2. Cracking Codes with AI: There's some computer brainpower trying to pick locks on the secret codes we already got checking if they can stand the heat.

3. Keeping Things Private but Making Them Smart: There's this thing called homomorphic encryption that lets AI work on stuff without seeing what's inside. Keeps it hush-hush while the AI does its thing (Gentry 2009).

The new stuff mixes AI's flex to adjust with the secure-info stuff of cryptography. This duo comes up with fresh ways to keep data safe and spot weak spots[18].

LEGAL PRECEDENTS IN AI PATENTS

The most relevant legal precedents regarding artificial intelligence patents are generally from the application of "Section 3(k)" of the Indian Patents Act, which excludes "mathematical methods, business methods, computer programs per se, and algorithms" from patentable subject matter, thus thwarting the prospect of purely software-based artificial intelligence products being patented. However, innovations concerning AI may qualify for patentability under CRI guidelines for patentability if they show a technical advancement or resolution of a technical problem.

Increasingly, as we link ourselves in the web of latter-day modernity, India and other countries that now attract patent expansions concerning CRI (computer-related invention) and AI must have put the issue on the dais of tax offices reviewing their respective patentability requirement in software. The reason is that these inventions are the excluded subject matter

The most basic level of software code commands the AI system to act, deduce, and produce an output based on pre-existing precepts like "if this condition is true, then perform the following action". Hence the more crucial question is how to vet patent filings for inventions pertaining to CRI and AI. Like any inventions, fundamental legal requirements of novelty, inventive step, and industrial application need to be satisfied by CRI and AI inventions, which are otherwise patentable. The IPO does not define what an invention is positively; rather it describes a non-exhaustive list of "non-inventions", defining subject matter and activities which are excluded from patentability for the extent of its relation as per Section 3(k) and 3(m) of the Indian Patents Act, 1970[19].

LEGAL AND ETHICAL ISSUES

Data Privacy:

 AI-related innovations in cryptography have to do with sensitive information. Therefore, it becomes extremely important to balance the security of intellectual property with the privacy and security in AI applications[20].

Ownership:

It raises the question of ownership, in situations where AI has contributed to making the cryptographic invention. If an AI system is used to invent some new algorithm in cryptography, then whether the inventor is the AI itself or its creator or the user of the AI.

International Patent Protection:

By their nature, cryptographic technology typically stretches across borders. Protecting the internationally based invention is complicated if considered involving AI-based cryptography and patent laws of the different countries with regard to all these patents and their regulations.

Data Protection:

The innovations in cryptography attributed to AI often correlate with sensitive information. Therefore, the intellectual property rights balancing aspect concerning privacy and security algorithm in AI applications is especially important[21].

GLOBAL PATENT LAWS AND AI

United States:

 AI-related inventions may be patented by the U.S. Patent and Trademark Office (USPTO), but the patent application must show that the AI or AI-based system is a component of a technical solution to a technical problem. AI cannot be the inventor, according to the USPTO (USPTO, 2020), a claim that is presently being debated in relation to inventions produced by AI[22].

European Union:

 If an AI-related invention solves a technical problem and produces a technical impact, it may be eligible for patent protection at the European Patent Office (EPO). However, mathematical techniques or abstract concepts—like some types of machine learning algorithms—cannot be patented. According to EPO (2020), AI cannot be credited as the inventor[23].

CASE LAWS

Alice Corp. V. CLS Bank (134 S. Ct. 2347 (2014))

In this case we can observe that the supreme court unanimously ruled the Alice Corporation’s patents claiming a computer-implemented method for managing financial risk through a neutral third party were invalid because they essentially claimed an abstract idea (a fundamental economic practice) without a sufficient inventive concept, meaning simply implementing the idea on a computer was not enough to make it patentable; this decision significantly impacted the patentability of software-based inventions[24].

Google LLC v. Oracle America Inc. (No. 18–956. Argued October 7, 2020—Decided April 5, 2021)

In the case of Google LLC versus Oracle America Inc., the US Supreme Court settled the dispute in a vote of 6-2 that Google's copying of an API had allowed for "fair use" so as not to constitute copyright infringement and thus gave Google's Android operating system an in-road to the Java API; henceforth, Google would not have to pay Oracle for any licensing fees, as it was a transformative use of the copied code for the development of a new platform for mobile devices; this case offers an important legal precedent concerning copyright protection for APIs in the software industry.[25]

CHALLENGES

Although AI cryptography offers many advantages in the context of security and efficiency of work done in it, addressing these challenges calls for rigorous teamwork to increase the realization of AI cryptography's potentials. Therefore, the researchers, developers, and policy manufacturers have to work cooperatively so that AI cryptographic systems can be robust, resource-efficient, privacy-preservation-oriented, ethically sound, and well-scaled for the changing requirements in areas such as secure communications and data protection. This will be elaborated further:

1.       Adversarial Attacks:

Adversarial attacks on the AI cryptography systems are a major concern. Malicious actors could employ adversarial machine learning techniques to manipulate AI models and compromise the security of encrypted data.

2.       Resource Requirements: 

The AI algorithms, especially the deep learning models, require significant computational resources combined with large amounts of data for training. Implementation of AI cryptography on resource-constrained devices or networks can be challenging.

3.       Ethical Considerations:

 Ethical issues regarding AI cryptography are quite complicated. The challenge is to strike a perfect balance between privacy and utility. Individual privacy rights must be properly protected within AI cryptographic systems, yet must also reach an understanding with respect to the security goals.

4.       Scalability: 

Most AI cryptographic algorithms are computationally expensive and a barrier to wide adoption, especially in real-time applications or systems that manipulate large volumes of data. Research should aim for the development of scalable AI cryptography capabilities that rely upon such techniques that do not compromise the security along with meeting high workloads[26].

CONCLUSION

It appears that there exists enormous promise in enhancing cryptographic security and patent protection by deploying AI systems. AI might develop highly secure algorithms for encryption, detect vulnerabilities of systems faster, and mount an agile defense against imminent threats in real-time. The integration of AI within the ambit of cryptography under patent law is beset with very serious difficulties, notably in terms of the intellectual property rights concerning the tension between innovation and the protection of proprietary technology.

Future recommendations include fostering cooperation between AI researchers, cryptographers, and patent offices to develop a framework that does not cease innovation but also protects intellectual property. More study should be advanced on AI-influenced cryptographic methods and patenting to seek new, more robust, and ever-evolving encryption paradigms providing solid defense against oncoming cyber threats.

[1]  anonymous, cryptography introduction, (geeksforgeeks 20, mar,2023), <https://www.geeksforgeeks.org/cryptography-introduction/>, accessed in 11th FEB 2025

[2]  anonymous, History of Cryptography, (IBM  05, JAN,2024), <https://www.ibm.com/think/topics/cryptography-history>, accessed in 12th Feb 2025

[3]anonymous, History of Cryptography, (IBM  05, JAN,2024), <https://www.ibm.com/think/topics/cryptography-history>, accessed in 12th Feb 2025

[4]  anonymous, History of Cryptography, (IBM  05, JAN,2024), <https://www.ibm.com/think/topics/cryptography-history>, accessed in 12th Feb 2025

[5] anonymous, History of Cryptography, (IBM  05, JAN,2024), <https://www.ibm.com/think/topics/cryptography-history>, accessed in 12th Feb 2025

[6] Corinne Bernstein, Definition of Cryptosystem,(TechTarget), <https://www.techtarget.com/searchsecurity/definition/cryptosystem#:~:text=The%20term%20cryptosystem%20is%20shorthand,can%20read%20and%20process%20it>, accessed in 13th FEB 2025

[7]Praneetha Neelapareddigari, AI in Cryptography, (Insights2Techinfo), < https://insights2techinfo.com/ai-in-cryptography/>, accessed in 13th FEB 2025

[8] Gerg Vetter,  Patent Protection, (Scholarship.kentlaw ,June,2009), <https://scholarship.kentlaw.iit.edu/cgi/viewcontent.cgi?article=3715&context=cklawreview>, accessed in 14th FEB, 2025

[9]Anonymous, Application of AI in Cryptography, (BSI), <https://www.bsi.bund.de/EN/Themen/Unternehmen-und-Organisationen/Informationen-und-Empfehlungen/Kryptografie/KI-in-der-Kryptografie/ki-in-der-kryptografie_node.html#:~:text=AI%20can%20be%20used%20to,preserving%20machine%20learning%20and%20inference>, accessed in 14th FEB

[10] Praneetha Neelapareddigari, AI in Cryptography, (Insights2Techinfo), < https://insights2techinfo.com/ai-in-cryptography/>, accessed in 13th FEB 2025

[11] Author: anonymous, Published by Springer Open, https://jast-journal.springeropen.com/articles/10.1186/s40543-024-00416-6#:~:text=Overview%20of%20AI%20techniques%20in,which%20helps%20enhance%20security%20measures.

[12] Anonymous, AI algorithms: a complete overview, (tableau), <https://www.tableau.com/data-insights/ai/algorithms#:~:text=At%20the%20core%20level%2C%20an,training%20data%20as%20a%20basis>, accessed in 13th FEB 2025

[13] Ray Kurzweil, Cryptography enhancing security and privacy in digital age, (Medium), <https://medium.com/@singularitynetambassadors/ai-cryptography-enhancing-security-and-privacy-in-the-digital-age-db5c1bbf5fdb#:~:text=AI%20cryptography%20enables%20the%20development,need%20to%20decrypt%20it%20first>, accessed in 15th FEB 2025

[14] Ismail Assoujaa, AI usage in Cryptography, (research gate October, 2024), <https://www.researchgate.net/publication/384968533_Artificial_Intelligence_usage_in_Cryptography#:~:text=The%20integration%20of%20artificial%20intelligence%20(AI)%20with,and%20enabling%20dynamic%20adaptation%20to%20evolving%20threats>, accessed in 15th FEB 2025

[15] Ray Kurzweil, Cryptography enhancing security and privacy in digital age, (Medium), <https://medium.com/@singularitynetambassadors/ai-cryptography-enhancing-security-and-privacy-in-the-digital-age-db5c1bbf5fdb#:~:text=AI%20cryptography%20enables%20the%20development,need%20to%20decrypt%20it%20first>, accessed in 15th FEB 2025

[16]Sheldon Brown, How to Patent an AI or Machine Learning Invention, (Patent Experts 30, Jan, 2024), <https://patentexperts.org/patent/how-to-get-a-patent/ai-patents/>, accessed in 14th Feb 2025

[17] Anonymous, AI in Patents, (Bit Law), <https://www.bitlaw.com/ai/AI-patents.html#:~:text=As%20is%20true%20for%20other,describe%20the%20issue%20at%20all>, accessed in 14th FEB 2025

[18]David Cain, Innovations in Cryptography and security using AI, (LinkedIn, 22, JULY, 2024) <https://www.linkedin.com/pulse/innovations-cryptography-security-using-ai-david-cain-ij9dc/>, accessed in 14th FEB 2025

[19] Shraddha Singh Chauhan, India: patent eligibility of AI- related inventions, (Managing IP 29, Sep, 2023), <https://www.managingip.com/article/2c8q22uq5bhkzgreo35s0/sponsored-content/india-patent-eligibility-of-ai-related-inventions#:~:text=IPD%2DPAT)%2029/2022,effect%20on%20the%20underlying%20software.%E2%80%9D>, accessed in 15th FEB 2025

[20] Anonymous, Understanding the Role of AI in online privacy & Data Protection, (ESET 20, March,2024), <https://www.eset.com/za/about/newsroom/press-releases-za/blog/understanding-the-role-of-ai-in-online-privacy-data-protection/#:~:text=Methods%20Used%20for%20AI%20Privacy,analysis%20and%20machine%20learning%20capabilities>, accessed in 16th FEB 2025

[21]Anonymous, Understanding the Role of AI in online privacy & Data Protection, (ESET 20, March,2024), <https://www.eset.com/za/about/newsroom/press-releases-za/blog/understanding-the-role-of-ai-in-online-privacy-data-protection/#:~:text=Methods%20Used%20for%20AI%20Privacy,analysis%20and%20machine%20learning%20capabilities>, accessed in 16th FEB 2025

[22] Anonymous, AI and Emerging Technology Partnership engagement and events, (uspto), <https://www.uspto.gov/initiatives/artificial-intelligence/ai-and-emerging-technology-partnership-engagement-and-events>, accessed in 16th FEB 2025

[23] Anonymous, European Patent Register, (epo), <https://www.epo.org/en/searching-for-patents/legal/register>, accessed in 16th FEB 2025

[24] Anonymous, Dissecting the Alice Case: 6 Years in the making, (Lexis Nexis, 23, June, 2020), < https://www.lexisnexisip.com/resources/dissecting-the-alice-case/#:~:text=In%20the%202014%20Alice%20case,computation%20or%20other%20general%20principle)>, accessed in 17th FEB 2025

[25] Anonymous, Google LLC v. Oracle America Inc., (Oyez), <https://www.oyez.org/cases/2020/18-956>, accesed  in 17th FEB 2025

[26] Ray Kurzweil, Cryptography enhancing security and privacy in digital age, (Medium), <https://medium.com/@singularitynetambassadors/ai-cryptography-enhancing-security-and-privacy-in-the-digital-age-db5c1bbf5fdb#:~:text=AI%20cryptography%20enables%20the%20development,need%20to%20decrypt%20it%20first>, accessed in 15th FEB 2025