Method for constructing double-anchored identity of personal AI agent combining autonomy and compliance

Abstract

This invention discloses a method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance. First, a hash chain structure is used to cryptographically bind the agent's executable program code to its owner's identity, generating a unique cryptographic commitment with integrity verification. Second, a two-factor distributed identifier is designed, which, through the fusion of user anchoring factors and multiple regulatory factors, supports conditional identity disclosure under regulatory authorization, ensuring that the agent's behavior complies with requirements. Then, a chameleon hash protocol is used to construct a dual-anchored personality credential, unifying the verified user identity with the agent's capability traceability, establishing a persistent root of trust for compliance while maintaining user data control. Finally, a secure lifecycle management mechanism ensures that the binding relationship between the agent and the user remains consistent throughout the agent's version evolution and credential recovery process, achieving autonomy maintenance and compliance management throughout the agent's entire lifecycle.

Description

Technical Field

[0001] This invention relates to the fields of digital identity management and artificial intelligence security technology, specifically to a method for constructing and cross-domain accountability of a trusted personal AI agent with dual anchored identity that integrates autonomy and compliance in an intelligent agent internet environment. Background Technology

[0002] In today's era of rapid development in digital and intelligent technologies, the Internet of Agents (IoA) foreshadows a new form of the Internet characterized by highly autonomous agents engaging in cross-domain collaboration, task orchestration, and value exchange. In this evolution, the Personal Agent (PA), as a digital surrogate of a natural person in a distributed network, can perceive and execute complex tasks across heterogeneous platform boundaries. Therefore, a unique, verifiable, and traceable agent identity becomes the anchor of trust for the orderly development of the Internet of Agents. Currently, Sovereign Identity (SSI) schemes are considered promising for building a foundation of trust. However, existing schemes fail to fully consider the strong semantic binding between the agent's entity attributes (such as code logic and configuration information) and the user's identity. This results in a lack of underlying intrinsic connection between the user's autonomy over the agent (i.e., the right to autonomously manage their identity and associated data), making the identity highly susceptible to cloning or illegal copying. Meanwhile, while existing solutions offer a degree of anonymity, they often rely on a single centralized authority to hold individuals accountable for malicious acts. This not only risks privacy breaches or internal misuse but also fails to meet the complex compliance requirements of IoA cross-domain interactions (i.e., compliance with laws and regulations, interaction protocols, and the ability to be audited by regulators). Furthermore, the lack of a comprehensive lifecycle management mechanism covering the entire lifecycle of an agent, from instantiation and capability evolution to behavioral tracing, makes it difficult to ensure the consistency of identity control. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance.

[0004] The method of this invention includes five entities: a user and its personal agent PA, a trust provider TP, an agent provider AP, and the InterPlanetary File System (IPFS); The method of this invention is divided into four stages: The first stage initializes the user, trust provider, and intelligent agent provider in the system, generating their respective encryption key pairs; the second stage involves the user authenticating themselves through the trust provider, obtaining basic personality credentials, and establishing their unique identifier in the physical world; the third stage involves the intelligent agent provider generating a cryptographic commitment for the user to trace their personal intelligent agent capabilities, and using a hash chain structure to immutably bind it to the user's identity; the fourth stage involves the trust provider and the intelligent agent provider collaborating through a security protocol to generate regulatory factors, which are then directly embedded into the personality credentials. The user uses the personality credentials to interact with privacy-preserving data through a two-factor DID, and, when necessary, supports conditional identity tracing by regulatory agencies to achieve compliance accountability.

[0005] Furthermore, user authentication and personality credential initialization are as follows: User-generated unique anchoring factor and random binding value and encapsulate it as an encrypted credential request. The request is sent to the trust provider TP; after verifying the validity of the request, the trust provider calculates the intermediate regulatory factor. It also uses a customized chameleon hash algorithm to bind intermediate regulatory factors with anchoring factors; simultaneously, it constructs a distributed identifier for users based on verification information. The trust provider generates a Personality Certificate (PHC) and its corresponding authorization context. and the encrypted response result Return to the user; the user decrypts the response and verifies the integrity of the binding relationship between the regulatory factor and the anchor factor.

[0006] Furthermore, anchoring factor Request for credentials The calculation is as follows:

[0007]

[0008] Where H represents the hash function, The encryption algorithm indicates the use of a public key. Encrypt; This is a signature algorithm, indicating the use of the user's private key. Sign up. This indicates that the binding factor is randomly generated by the user, BI represents the user's biometric identifier, and PII represents the user's generated personal identity information.

[0009] Furthermore, firstly, the trust provider receives the credential request. Then, using its private key Decryption:

[0010] in, This indicates the use of TP's private key. The algorithm used for decryption; Next, verify the embedded signature. and anchoring factor The correctness; After successful verification, TP uses its Paillier public key. Homomorphic encryption is performed on the identity hash value to calculate the intermediate state regulatory factor. :

[0011] in, A value randomly selected for TP; Then, TP utilizes a custom chameleon hash. The algorithm will Bind to In the process, new random values ​​are generated. : ;in Indicates A custom chameleon hash algorithm is executed as the trapdoor key; the anchor factor is updated to: ,in Indicates will Binding middle; After binding is complete, TP uses the key derivation function. Calculate the encryption key and utilize Encryption generates secure information packets It was then uploaded to the InterPlanetary File System (IPFS) to obtain a content identifier. ; Calculate the encrypted content identifier using the Paillier encryption algorithm :

[0012] in, This refers to the Paillier encryption algorithm, used in Paillier cryptography. Perform encryption operations on encrypted data.

[0013] Furthermore, in order to issue users their exclusive personality credentials, the trust provider TP first initializes a personality credential template. This template is a structured instance with all fields initially set to empty, and logically consists of the following four parts. (1) Data owned by the agent itself: a composite field containing identity data maintained by the agent; the composite field is divided into two subfields: TP management and issuance items and AP management and issuance items; (2) TP-related proof: a field used to record verification evidence issued by the trust provider; (3) AP-related proofs: fields used to record verification evidence issued by the agent provider; (4) Personality Certificate Verification: Security field, used to record all signatures of the personality certificate; Specifically, TP assigns attribute values ​​to the ASO.TPM field of the personality credential. and generate proof TP selects two random factors. And generate the Chameleon hash value of the trust provider based on the TP public key. Simultaneously, TP uses AP's public key to generate the chameleon hash value for the agent provider. Subsequently, a binding proof is generated by signing the concatenation of the two hash values ​​mentioned above. The TP fills in the relevant information into the corresponding fields of the personality credential; secondly, in order to achieve compliant updates of the PHC by the AP, the TP generates context information for subsequent authorization. :

[0014] TP uses identity hash value As an index, related parameters Store the encrypted final response in the local database. Send to the user.

[0015] Furthermore, the user received Then, using its private key Decryption: Subsequently, the user performs data extraction and verification checks; first, the user extracts anchoring factors from the Personality Certificate (PHC). And extract intermediate regulatory factors from the authorization context TPAC. Finally, users verify the anchoring factor using the following formula. Binding consistency: If all verifications pass, it means that the initial PHC has been successfully generated, and the anchoring of the user's real identity with the PHC has been completed.

[0016] Further, the generation of the user's personal intelligent agent is verified, and the specific steps are as follows: First, the user sends an encrypted agent request to apply for configuration of a personal agent. Then, the agent provider receives the user's identity credentials and verifies the binding relationship between the user's identity and anchoring factors, constructing and returning a custom module matrix accordingly. Next, the user decrypts the matrix and provides the selected configuration information. The agent provider generates a complete personal agent based on this selection and updates the identity credentials by modifying the hash chain, ensuring the immutability of the agent's technical traceability. Finally, after verifying the consistency of the agent, the user sends a generation request to the trust provider, which ultimately generates the user's distributed identifier and final identity credentials.

[0017] Further, the generation of the user's personal intelligent agent is verified, and the specific steps are as follows: STEPD1. The user uses the public key provided by the agent provider. For Personality Certificate (PHC) and User Identity Hash Value Intermediate regulatory factors and the trust provider's authorization context Encryption is performed to construct the agent's request. And send it to the agent provider AP; STEPD2. The agent provider AP receives... Then, using its private key Decryption is performed; subsequently, the AP uses the identity hash value. Search the local database to ensure that there is no existing record for the user, in order to confirm the legitimacy of the request and initiate the verification process; AP extracts anchoring factors from personality credential PHC. With proof ; and from Extracting intermediate regulatory factors Updated binding random value And encrypted random factors The AP performs two key verifications: 1. Based on the TP public key, it verifies the signature of the TP proof and the authorization context using a preset verification algorithm; 2. It verifies the correctness of the binding factor. After verification, AP constructs a custom module matrix (CMM) as a configuration matrix containing options for each functional module of the agent; Finally, the AP uses the user's public key. For the construction Encrypt and send the encryption result Return to the user; STEPD3. After receiving the custom module matrix encrypted by the AP, the user uses its private key. Decrypt to obtain optional information about the module: ,in This indicates a decryption algorithm used based on the user's private key. Perform decryption on encrypted data; Users exercise the right to define the logical architecture of the intelligent agent, select target modules from a custom module matrix according to actual needs, and generate vectors representing the indices and descriptions of the selected modules: ,in Indicates the first The index of the specific modules selected in each functional category; Finally, the user utilizes the public key provided by the agent provider. For the selected custom module combination Encryption is performed to construct customized information for the personal intelligent agent. and return it to AP: ; STEPD4. The agent provider (AP) uses its private key. The received encrypted combination of custom modules is decrypted to retrieve the module index; then the corresponding code block is located based on the index, and the complete custom module code is reconstructed. AP executes personalized agent generation method This method generates a complete Personal Agent (PA) using a verified Personality Certificate (PHC) and retrieved code modules. As input, the target functional components are reorganized and encapsulated according to a preset logical topology to instantiate a functional autonomous intelligent agent that is cryptographically strongly bound to the user's personality attributes. AP will customize module code With identity hash value Combined, through the hash chain generation algorithm ( Construct configuration commitment ; AP uses Algorithm I to update PHC, anchoring the agent's configuration code to PHC; after updating the personality credential PHC, AP uses a pseudo-random function to generate intermediate variables. Subsequently, combined with intermediate regulatory factors The final regulatory factor is calculated using the Paillier homomorphic encryption algorithm. Ultimately, this enables compliant identity traceability for intelligent agents; AP uses identity hash value As an index, a set of key parameters is stored in the local database, including This is to provide a basis for subsequent auditing and maintenance; AP assigns domain name identifiers to users and using the user's public key The updated Personality Certificate (PHC) and regulatory factors Domain name identifier The final response result is constructed by encrypting a personalized smart PA containing randomly bound values. : ; STEPD5. The user receives the final response from the personal intelligent agent. and use its private key Decrypt to obtain the updated binding random value. Regulatory factors Domain name identifier Ultimate proof of personality And personal intelligent agent (PA): Subsequently, the user extracts the custom module code corresponding to each module from PA. : ; Users utilize the extracted custom module code and identity hash value Recalculate configuration commitment Users retrieve stored information from their Personality Certificate (PHC). and Perform a consistency check; the user performs final verification to ensure the configuration commitments are fulfilled. Correctly bound to anchor factor middle; If all the above verifications pass, the user can utilize... The regulatory factor, domain name identifier, and updated credentials are encrypted, and a distributed identifier generation request is constructed and sent to the trust provider. ; STEPD6. Trust providers use their private keys Received distributed identifier generation request Decrypt to obtain the relevant content: ; TP constructs self-certification identifier and use its public key Calculate the corresponding encrypted self-authentication identifier ; Subsequently, TP executes Algorithm II to generate the final Personality Certificate (PHC)'' and sends it to the user. If the user successfully receives and stores the PHC'', the user obtains a unique dual-anchored identity PHC.

[0018] Furthermore, Algorithm I is an AP-driven PHC update algorithm, specifically implemented as follows: Algorithm I input: Module commitment Identity hash Personality certificate to be modified AP Modification Factor and binding factor ; Algorithm I is executed as follows: Agent credential generation: ; ; ; ; Chameleon Collision Hash Search: ; Search for results that meet the following conditions : ; use Algorithm to obtain : ; Update anchoring factor : ; Algorithm I Output: .

[0019] Furthermore, Algorithm II is a TP-driven PHC update algorithm, specifically implemented as follows: The input to Algorithm II is: distributed identifier. Personality certificate to be modified and TP modification factor ; Algorithm II is executed as follows: TP proof generation: ; ; ; ; Local data query : ; Search for results that meet the following conditions : , ; Algorithm II output: .

[0020] The beneficial effects of this invention are as follows: This invention implements an integrity verification and individual uniqueness guarantee mechanism for personal intelligent agents. By introducing hash chain technology, the executable code of the personal agent (PA) is cryptographically bound to the user's identity information to generate a unique verification identifier, which provides an immutable integrity commitment to the internal configuration of the agent. Compared with existing methods of generating agents based on shared code library instantiation, the personal intelligent agent constructed by this invention possesses non-replicable and anti-cloning characteristics, thereby effectively preventing the agent's identity from being illegally impersonated and its operational logic from being maliciously tampered with.

[0021] This invention constructs a dual-anchored trust root mechanism. By introducing Personality Credentials (PHC) as the core trust carrier and combining it with a customized Chameleon Hash protocol, it achieves synchronous binding between the user's decentralized identity (DID) and the agent's capability traceability identifier. This mechanism ensures an inseparable association between the user's identity and the agent's permissions, thereby effectively resisting identity forgery attacks and the risk of capability source tampering, and improving the overall trustworthiness and consistency of the system.

[0022] This invention proposes a dynamic balance mechanism that balances privacy protection and regulatory compliance. By constructing a two-factor decentralized identity (DID) system, a user-controlled anchor factor (AF) is jointly bound to a regulatory factor (RF) generated based on biometric recognition. Under normal business scenarios, users can interact anonymously using AF; when specific compliance conditions are met, a collaborative authorization mechanism between the trust provider (TP) and the agent provider (AP) enables controllable disclosure and traceability of identity, thereby addressing the shortcomings of existing decentralized identity systems in terms of conditional regulation.

[0023] This invention further provides a flexible and secure system-level lifecycle management method. By designing a controlled state transition mechanism, it achieves a secure transition for personal intelligent agents during configuration updates, version evolutions, and instance replacements, while ensuring that the root of trust is not compromised. Simultaneously, based on a trusted party verification-based identity recovery mechanism, it supports identity reconstruction and binding relationship restoration in the event of lost or abnormal personal credentials, thereby avoiding identity unavailability issues caused by single-point credential failure and improving system stability and fault recovery capabilities.

[0024] In summary, this invention, by constructing a personality credential system with dual anchoring features, achieves a unified approach to ensuring the uniqueness of an individual's identity, protecting privacy, and maintaining regulatory controllability in the intelligent agent internet environment, as well as providing secure management throughout the entire lifecycle. It possesses excellent security, scalability, and practical application value. Attached Figure Description

[0025] Figure 1 A schematic diagram of the system model of this invention.

[0026] Figure 2 A schematic diagram of the webpage system of the present invention.

[0027] Figure 3 This invention provides a schematic diagram of cross-domain access based on the ANP protocol. Detailed Implementation

[0028] This invention proposes a dual-anchored identity construction method for personal AI agents that integrates autonomy and compliance, aiming to achieve a balance between deep autonomy of personal AI agents and compliant system supervision. First, a hash chain structure is used to cryptographically bind the executable program code of the AI ​​agent to its owner's identity, generating a unique cryptographic commitment with integrity verification, thus preventing cloning and tampering. Second, a two-factor distributed identifier (DID) is designed, integrating user anchoring factors with multiple regulatory factors to support conditional identity disclosure under regulatory authorization, ensuring that AI agent behavior complies with IoA interaction rules and legal compliance requirements, while guaranteeing users' independent control over their identity information. Furthermore, a dual-anchored personality credential (PHC) is constructed using the Chameleon hash protocol, unifying the verified user identity and the traceability of AI agent capabilities, establishing a persistent root of trust for compliance while maintaining user data control. Finally, a secure lifecycle management mechanism ensures that the binding relationship between the AI ​​agent and the user remains consistent throughout version evolution and credential recovery, achieving autonomy maintenance and compliant management throughout the entire lifecycle of the AI ​​agent.

[0029] Parameter description:

[0030] The method of this invention mainly includes five entities, such as Figure 1 As shown, it includes: a user and their personal agent (PA), a trust provider (TP), an agent provider (AP), and the InterPlanetary File System (IPFS).

[0031] User: Refers to the physical individual seeking a customized personal agent. The user authenticates with a trusted provider to obtain a Personality Certificate (PHC), and then submits the certificate to the agent provider to trigger the configuration and deployment of the Personal Agent (PA).

[0032] Personal Agent (PA): A dedicated intelligent agent uniquely coupled to a single user. In the digital ecosystem, the PA acts as the user's autonomous representative, performing tasks and making decisions based on specific authorization policies and personality parameters defined during the creation phase.

[0033] Trust Provider (TP): As the core institution for identity assurance, the Trust Provider verifies user identity through physical world authentication methods such as biometric information and government-issued digital identities. Upon successful verification, it issues a Personality Certificate (PHC) as cryptographic proof of the user's unique and authentic identity.

[0034] Agent Provider (AP): Refers to the organizational entity that provides agent services to users. The agent provider executes a secure generation protocol, produces customized personal agents (PAs) according to user needs, and ensures that the agent is cryptographically and logically strongly bound to the personality credential (PHC) submitted by the user.

[0035] InterPlanetary File System (IPFS): A decentralized storage layer for encrypted user data. IPFS ensures data availability and integrity, with data confidentiality guaranteed by pre-defined encryption mechanisms.

[0036] The method of this invention mainly consists of four stages. The first stage initializes various entities in the system, including the user, trust provider (TP), and agent provider (AP), generating their respective encryption key pairs to establish a foundational trust root for subsequent cross-domain collaboration and secure communication. The second stage involves the user authenticating themselves through the trust provider, obtaining basic identity credentials, and establishing their unique identifier in the physical world, such as... Figure 1 Steps 1-3 are shown. In the third stage, the agent provider generates a cryptographic commitment for the user to trace their personal agent capabilities and uses a hash chain structure to immutably bind it to the user's identity, ensuring the autonomy of the personal agent from the bottom layer and preventing the identity from being illegally cloned or tampered with, such as... Figure 1 Steps 4-8 are shown above. In the fourth stage, the trust provider and the agent provider collaborate through a security protocol to generate regulatory factors, which are then directly embedded into the identity credential. Users then use this credential for privacy-protected interactions via two-factor ID (DID), and, when necessary, support conditional identity tracing by regulatory agencies to achieve compliance accountability. Figure 1 Step 9 is shown.

[0037] I. Initialization Phase First, the trust provider (TP) performs initialization to generate public parameters. Specifically, it first selects a prime number. and cyclic groups Generator ; an elliptic curve Its order is a prime number. The generator is Select an RSA module Select Generators in space Used for Paillier encryption. Then, it will use public parameters. Released to other entities.

[0038] After receiving the payment, the user selects the private key. And calculate the corresponding public key based on the discrete logarithm (DL) group: Users securely store their private keys. and publish the public key .

[0039] The agent provider (AP) selects its private key. And calculate the corresponding public key set: AP securely stores private keys and publish public keys .

[0040] Trust Provider (TP) selects its private key And calculate the corresponding public key set: In addition, TP generates the Paillier public key. and the corresponding Paillier private key TP securely stores private keys. and and publish public keys , as well as .

[0041] In the Paillier encryption algorithm, given plaintext... and a random parameter The ciphertext c is calculated as shown in equation (1).

[0042]

[0043] For the Paillier decryption algorithm, given the ciphertext... Calculate plaintext As shown in equation (2):

[0044] II. User authentication and Personality Certificate (PHC) initialization; User-generated unique anchoring factor and random binding value and encapsulate it as an encrypted credential request. The request is sent to the Trust Provider (TP). After verifying the validity of the request, the Trust Provider calculates the intermediate regulatory factor. And a customized chameleon hash algorithm is used to bind intermediate regulatory factors with anchoring factors; at the same time, a distributed identifier for the user is constructed based on the verification information. The trust provider generates a Personality Certificate (PHC) and its corresponding authorization context. and the encrypted response result The system returns to the user. The user decrypts the response and verifies the integrity of the binding relationship between the regulatory factor and the anchoring factor. The specific process is as follows: STEPG1. Users first generate their own biometric identifier (BI), such as iris and fingerprint data; then, users generate their Personal Identification Information (PII), including their ID card number, phone number, and name. Next, users randomly generate binding factors. ; and utilize the public key of the agent provider. and the public key of the trust provider The anchoring factor is obtained through calculation. anchoring factor As a technological representation of the user's right to exercise autonomy, it enables autonomous control over the identity of the intelligent agent. The specific calculation is shown in equation (3):

[0045] Where H represents a hash function.

[0046] Users use the public key of the trusted provider. For the calculated The public key of the smart agent provider The credential request is constructed by encrypting the biometric identifier (BI), personally identifiable information (PII), and the signature for the binding factor. As shown in equation (4).

[0047]

[0048] in, The encryption algorithm indicates the use of a public key. Encrypt it. This is a signature algorithm, indicating the use of the user's private key. Sign it.

[0049] Finally, the constructed credential request will be submitted. Send to trusted provider TP.

[0050] STEPG2. Trust Provider (TP) receives credential request. Then, using its private key Decryption:

[0051] in This indicates the use of TP's private key. The algorithm used for decryption.

[0052] And verify the embedded signature. and anchoring factor Correctness: Whether it is valid or not.

[0053] After successful verification, TP uses its Paillier public key. Homomorphic encryption is performed on the identity hash value, and an intermediate regulatory factor is calculated to achieve compliance with regulations. As shown in equation (5).

[0054]

[0055] in, A value randomly selected for TP.

[0056] Then, TP used a custom chameleon hash ( The algorithm will Bind to In the process, new random values ​​are generated. : .in Indicates A custom chameleon hash algorithm executed as a trapdoor key.

[0057] Accordingly, the anchoring factor is updated as follows: ,in Indicates will Binding In the middle. At this time. The system internally binds a new intermediate regulatory factor, but its hash result remains unchanged, thereby achieving the endogenous embedding of compliance information without altering the user's initial identity, ensuring the user's autonomous control over their identity.

[0058] After binding is complete, TP uses the key derivation function ( Calculate the encryption key and utilize Encrypted biometrics (BI), personally identifiable information (PII), and other data generate secure information packets. ); then upload it to the InterPlanetary File System (IPFS) to obtain a content identifier. To ensure privacy and integrity, the Paillier encryption algorithm is used to calculate the encrypted content identifier. As shown in equation (6):

[0059] in, This refers to the Paillier encryption algorithm, used in Paillier cryptography. Perform encryption operations on encrypted data.

[0060] STEPG3. To issue a user's unique personality credential, the Trust Provider (TP) first initializes a personality credential template. This template is a structured instance, with all fields initially set to null. Its detailed construction is shown in Table 1, and logically consists of the following four parts: (1) Agent-owned data (ASO): A composite field containing identity data maintained by the agent, further divided into two subfields: TP Management Issued Item (TPM): Privileged fields that can only be modified by the trust provider. AP Management Issued Item (APM): Context-dependent fields that can be modified by the agent provider (AP) after authorization by the trust provider.

[0061] (2) TP-related proof (TPA): This field is used to record verification evidence issued by the trust provider to endorse the integrity of the platform and the authenticity of the TPM.

[0062] (3) AP-related proof (APA): This field is used to record the verification evidence issued by the agent provider to endorse the integrity of the platform and the authenticity of APM.

[0063] (4) Proof of Personality (PROOF): A security field used to record all signatures of the personal identity certificate to achieve overall integrity protection.

[0064] Table 1

[0065] TP assigns attribute values ​​to the ASO.TPM field of the personality credential. and generate proof TP selects two random factors. And generate the Chameleon hash value of the trust provider based on the TP public key. Simultaneously, TP uses AP's public key to generate the chameleon hash value for the agent provider. Subsequently, a binding proof is generated by signing the concatenation of the two hash values ​​mentioned above. The TP then fills in the relevant information into the corresponding fields of the personality credential. Secondly, to enable the AP to update the PHC in compliance, the TP generates context information for subsequent authorization. As shown in equation (7):

[0066] TP uses identity hash value As an index, related parameters Store the encrypted final response in the local database. Send to the user.

[0067] STEPG4. When the user receives Then, using its private key Decryption: Subsequently, the user performs data extraction and verification checks. First, the user extracts anchoring factors from the Personality Certificate (PHC). And extract intermediate regulatory factors from the authorization context TPAC. Finally, users verify the anchoring factor using the following formula. Binding consistency: If all verifications pass, it means that the initial PHC has been successfully generated, and the anchoring of the user's real identity with the PHC has been completed.

[0068] III. Generation of Personal Intelligent Agents for Verified Users This phase mainly includes the following steps: First, the user sends an encrypted agent request to apply for the configuration of a personal agent (PA); then, the agent provider (AP) receives the personality credential (PHC) presented by the user and verifies the binding relationship between the user's identity and the anchoring factor, and constructs and returns a custom module matrix accordingly; next, the user decrypts the matrix and provides feedback on the selected configuration information, and the agent provider generates a complete personal agent based on the selection, and updates the personality credential by modifying the hash chain to ensure the immutability of the agent's technical traceability; finally, after verifying the consistency of the agent, the user sends a generation request to the trust provider, which ultimately generates the user's distributed identifier (DID) and final personality credential.

[0069] STEPD1. The user uses the public key provided by the agent provider. For Personality Certificate (PHC) and User Identity Hash Value Intermediate regulatory factors and the trust provider's authorization context Encryption is performed to construct the agent's request. And send it to the agent provider AP:

[0070] STEPD2. The agent provider AP receives... Then, using its private key Decryption: Subsequently, the AP used the identity hash value Search the local database to ensure that no existing record for the user exists, in order to verify the legitimacy of the request and initiate the verification process.

[0071] AP extracts anchoring factors from personality credential PHC. With proof ; and from Extracting intermediate regulatory factors Updated binding random value And encrypted random factors Next, the AP performs two key verifications: (1) Based on the TP public key, it verifies the signature of the TP proof and the authorization context using a preset signature verification algorithm: and (2) Verify the correctness of the binding factor: .

[0072] After verification, the AP constructs a custom module matrix (CMM), which serves as a configuration matrix containing options for each functional module of the agent.

[0073] in, The number of functional categories representing different modules. This represents the specific implementation method under each category.

[0074] Finally, the AP uses the user's public key. For the construction Encrypt and send the encryption result Return it to the user.

[0075] STEPD3. After receiving the custom module matrix encrypted by the AP, the user uses its private key. Decrypt to obtain optional information about the module: ,in This indicates a decryption algorithm used based on the user's private key. Perform decryption on encrypted data.

[0076] Users exercise the right to define the logical architecture of the intelligent agent, select target modules from a custom module matrix according to actual needs, and generate vectors representing the indices and descriptions of the selected modules: ,in Indicates the first The index of the specific module selected in each functional category.

[0077] Finally, the user utilizes the public key provided by the agent provider. For the selected custom module combination Encryption is performed to construct customized information for the personal intelligent agent. and return it to AP: .

[0078] STEPD4. The agent provider (AP) uses its private key. Decrypt the received encrypted combination of custom modules to retrieve the module index: Subsequently, the corresponding code block is located based on the index, and the complete custom module code is reconstructed. : .

[0079] AP executes personalized agent generation method Generate a complete personal intelligent agent (PA): This method uses verified Personality Certificates (PHCs) and retrieved code modules. As input, the target functional components are reorganized and encapsulated according to a preset logical topology to instantiate a functional autonomous intelligent agent that is cryptographically strongly bound to the user's personality attributes.

[0080] AP will customize module code With identity hash value Combined, through the hash chain generation algorithm ( Construct configuration commitment As shown in equation (8):

[0081] AP uses Algorithm 1 to update the PHC, thus anchoring the agent's configuration code to the PHC. After updating the Personality Certification (PHC), AP uses a pseudo-random function to generate intermediate variables. Subsequently, intermediate regulatory factors were incorporated. The final regulatory factor is calculated using the Paillier homomorphic encryption algorithm. Ultimately, this enables compliant identity tracing of intelligent agents. The specific calculation is shown in equation (9):

[0082] AP uses identity hash value As an index, a set of key parameters is stored in the local database, including This will provide a basis for subsequent auditing and maintenance.

[0083] AP assigns domain name identifiers to users and using the user's public key The updated Personality Certificate (PHC) and regulatory factors Domain name identifier The final response result is constructed by encrypting a personalized smart PA containing randomly bound values. :

[0084]

[0085] STEPD5. The user receives the final response from the personal intelligent agent. and use its private key Decrypt to obtain the updated binding random value. Regulatory factors Domain name identifier Ultimate proof of personality And personal intelligent agent (PA): Subsequently, the user extracts the custom module code corresponding to each module from PA. : .

[0086] Users utilize the extracted custom module code and identity hash value Recalculate configuration commitment : Users can further extract stored information from the Personality Certificate (PHC). And perform a consistency comparison: determine Whether it holds true. The user performs final verification to ensure the configuration commitment is valid. Correctly bound to anchor factor middle: , among which Indicates will Binding middle, Indicates will Binding middle.

[0087] If all the above verifications pass, the user can utilize... The regulatory factor, domain name identifier, and updated credentials are encrypted, and a distributed identifier generation request is constructed and sent to the trust provider. :

[0088] STEPD6. Trust Provider (TP) uses its private key Received distributed identifier generation request Decrypt to obtain the relevant content: .

[0089] TP constructs self-certification identifier and use its public key

[0090] Calculate the corresponding encrypted self-authentication identifier : TP defines the user's distributed identifier as... ,in, To identify specific identity methods in accordance with the W3C (World Wide Web Consortium) Decentralized Identifier (DID) technical specification. It is a domain name identifier assigned to the user by TP. It is a domain name identifier assigned by the AP user.

[0091] Subsequently, TP executes Algorithm 2 to generate the final Personality Certificate (PHC) and sends it to the user. If the user successfully receives and stores the PHC, the user obtains a unique dual-anchored identity (PHC).

[0092]

[0093] IV. Regulation of Personal Intelligent Agents Targeting Malicious Behavior When an individual agent engages in malicious behavior, the TP can disclose the agent's identity information based on the DID. First, the TP sends a request to the AP, utilizing regulatory factors. Without exposing plaintext information, reverse parse and obtain the user's identity hash value. TP utilizes the acquired Locate the corresponding storage identifier ( And retrieve encrypted records from the IPFS distributed storage system. Subsequently, the decryption key derived from the user's regulatory factors is used. The record is decrypted to reconstruct the user's biometrics. ) and personally identifiable information ( ).

[0094] After obtaining the Distributed Identifier (DID) of the malicious agent, STEPR1.TP extracts the Encrypted Secure Content Identifier (CSCID) from it and uses its private key. Decryption is performed to recover the anchoring factor. With regulatory factors : .

[0095] To achieve conditional oversight compliance, the TP also needs to send an oversight request to the AP, which includes parameter pairs. If the AP agrees to the regulatory request, a random factor is calculated based on local private parameters. And return it to TP.

[0096] TP is ultimately based on the received Using the Paillier homomorphic decryption process to analyze regulatory factors Decryption is performed to recover the user's identity hash value. :

[0097] STEPR2.TP is based on identity hash value Retrieve the corresponding encrypted content identifier from the local database. and using his Paillier private key Perform the decryption operation:

[0098] Subsequently, TP utilized the restored Encrypted security packets were retrieved from IPFS. TP uses its private key Regulatory factors and Paillier key parameters Calculation is performed using a key derivation function (KDF). :

[0099] Finally, using the derived key right Decryption is performed to extract the owner's biometric identifier (BI) and personally identifiable information (PII):

[0100] Thus, TP has completed the technical disclosure of the identity of the malicious agent's owner, under the premise of meeting the conditional regulatory compliance requirements, and achieved a closed loop of compliance and accountability in an autonomous environment.

[0101] V. Lifecycle Management of Personal Intelligent Agents 3.1 Recovery of the Personal Intelligent Agent This invention provides a method for restoring the identity of a verified user's Personal Agent (PA), aiming to achieve autonomous restoration of the PA by leveraging the user's possession and association of core credentials, thereby realizing continuous control over the PA. Its main features include the following steps: First, in a scenario where the user has lost their PA but possesses a valid Personal Identity Certificate (PHC), the user initiates an encrypted restoration request; subsequently, the Agent Provider (AP) verifies the integrity of the received credentials and, based on this, retrieves specific module code linked to the user's identity from its local database. Next, the AP uses the retrieved code to reconstruct and instantiate the complete personal intelligent agent, ensuring that its functional configuration is consistent with the original state. Finally, the AP encrypts the reconstructed personal intelligent agent and returns it to the user, who then performs decryption, consistency verification, and local secure storage, thereby achieving lossless migration and recovery of the personal intelligent agent's logic.

[0102] STEPRE1. The user initiates the restoration process of their Personal Agent (PA). The user utilizes the public key provided by the agent provider. The hash value of the user's identity identifier The Personal Data Identity (PHC) is encrypted to construct a recovery request. Send to AP:

[0103] STEPRE2.AP uses its private key For the received Decryption:

[0104] AP extracts anchoring factors from personality credential PHC. Configuration commitment and TP proof And Chameleon Hash Proof AP performs the following key verifications to ensure the legitimacy of the restore request: [Verification details to be filled in] and Verify the legitimacy of the PHC; utilize the identity hash value. Retrieve stored data from the local database. And compare the extracted from the voucher With local storage Are they consistent? Re-verify the anchoring factor Correctness: .

[0105] If all the above verifications pass, AP will regenerate the complete personal intelligent agent PA: Finally, the AP uses the user's public key. The reconstructed PA is encrypted, and the final response PAR is returned to the user: .

[0106] STEPRE3. The user receives the final response result PAR and uses their private key. Decryption is performed to recover the Personality Token (PHC) and Personal Agent (PA); subsequently, the user uses the PA to reconstruct the code blocks of each functional module. .

[0107] Users utilize the restored module code and identity hash value Recalculate code commitment value : Subsequently, the user retrieves the stored commitment from the Personality Certificate (PHC). And perform the final consistency check: .

[0108] If the above verification is successful, the Personal Intelligent Agent (PA) is deemed to have been successfully restored.

[0109] 3.2 Functional Updates to Personal Intelligent Agents This invention provides a method for updating a verified user's personal agent, aiming to dynamically control the evolution of the agent's capabilities and ensure the continuity of autonomy during the iterative process of agent functionality. The main steps include: First, the user initiates an encrypted update request to start the process; then, the agent provider (AP), after verifying the user's identity and anchor factor binding relationship, provides a custom module matrix as a configuration benchmark; next, the user decrypts the matrix and provides the selected specific configuration information; the AP then performs a system update, updating the binding relationship between the Personality Certificate (PHC) and the Personal Agent (PA) by synchronously modifying the hash chain, ensuring the continuity of technical traceability; finally, the user performs a comprehensive consistency verification of the update result, confirming that the cryptographic binding between the certificate and the agent is correct, and then completes local archiving.

[0110] STEPU1. The user initiates an update process for their Personal Agent (PA). The user utilizes the public key provided by the agent provider. PHC as well as Encryption is performed to construct the update request: The user will then update the request. Send to AP.

[0111] STEPU2.AP uses its private key For the received update request Decryption: Subsequently, AP extracted the anchoring factor from PHC. Configuration commitment and TP proof and The AP performs the same authentication and database retrieval process as the PA application phase. This includes verifying the validity of various credentials, retrieving local database records, and configuring commitments. Consistency verification and anchoring factors Binding relationship verification.

[0112] If all the above verifications pass, then AP constructs a custom module matrix. As a configuration baseline provided to users for selection:

[0113] in, Represents the number of functional categories. This represents the specific implementation method under each category.

[0114] Finally, the AP uses the user's public key. For the construction Encrypt and send the encryption result The information is then returned to the user. The user then needs to customize the complete PA information, a process that follows the same user module selection logic as the Personal Agent (PA) application phase. Specifically, the user first decrypts the custom module matrix. The system then selects a combination of target modules based on update requirements; finally, the user constructs encrypted personalized agent information using the same cryptographic processing procedures. And submit it to AP.

[0115] STEPU3.AP uses its private key Decrypt the received encrypted custom module information, locate the specific code block according to the index, and reconstruct the custom module code. :

[0116]

[0117] Subsequently, AP used the personal agent generation method to generate a complete and updated version of the personal agent: At the same time, the new custom module code will be added. With identity hash value Combined, generate the updated configuration commitment: Next, AP executes Algorithm 1 to update the PHC. Specifically, this is based on the updated configuration commitment. Generate new governance parameters using the current timestamp:

[0118] AP uses identity hash value As a retrieval index, the parameter set in its local database is updated and archived. Finally, the AP uses the user's public key. The updated personality credential PHC''', personal agent PA'', and random binding value Encrypt and construct the final response result. And return it to the user:

[0119] User received Verification is required, and this process follows the same user verification logic as the Personal Agent (PA) application phase. Specifically, the user comprehensively verifies the success of the updated PA generation and its cryptographic binding relationship with the Personality Certificate (PHC). Successful verification indicates that the PA information update was successful.

[0120] 3.3 Personality credential retrieval and personal intelligent agent reconstruction This invention provides a method for restoring credentials and agents based on user private information. It aims to restore the Personal Identity Certificate (PHC) and Personal Agent (PA) by leveraging the user's absolute control over their private information, thereby maintaining the continuity of identity and full lifecycle control within autonomy. The method is characterized by a layered recovery mechanism: First, the user initiates a restoration request to the Trust Provider (TP) by resubmitting their private identity information. The TP then uses Paillier homomorphic decryption technology and a Regulatory Factor (RF) to restore the encrypted index from distributed storage and decrypt it to obtain the original PHC. Next, the user initiates an agent restoration request to the Agent Provider (AP) using the restored PHC as the core binding factor. The AP then regenerates and instantiates the Personal Agent (PA) based on credential consistency verification. Finally, the user performs security verification and local archiving on the reconstructed agent, thus achieving complete reconstruction and binding from the underlying digital identity to the high-level execution logic.

[0121] STEPT1. The user uses the public key of the trusted provider. Its biometric information and personally identifiable information Encrypt the identifier (ID) and generate a restoration request: .

[0122] TP in acquisition After the request is made, operations consistent with regulatory procedures are performed to retrieve the encrypted record. Restore user's sensitive information Finally, TP performs a consistency check to determine whether the information submitted by the user matches the restored information:

[0123] If the verification passes, TP retrieves the user's initial PHC and uses the user's public key. Encrypt the restored PHC and then... Return it to the user.

[0124] STEPT2. After successfully obtaining the initial Personal Identity Certificate (PHC), the user initiates the recovery procedure of the Personal Agent (PA) by calling a preset restoration process (this process has been defined in 3.1). In this stage, the PHC serves as the core identity binding factor, and the AP regenerates the corresponding PA instance based on the consistency of the PHC. Finally, the user performs a final consistency verification and completes local secure storage after confirmation. At this point, the user has successfully achieved autonomous recovery of the PHC and PA, exercising the autonomous right to restore identity while ensuring that the restored agent identifier remains consistent with the original regulatory anchor, thus achieving a closed loop of autonomous control and compliance traceability.

[0125]

[0126] Those skilled in the art will understand that the embodiments in this application can be provided as methods, systems, or computer program products. Therefore, the embodiments in this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects.

Claims

1. A method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance, characterized in that: This method includes five entities: a user and their personal agent PA, a trust provider TP, an agent provider AP, and the InterPlanetary File System (IPFS); The method consists of four stages: The first stage initializes the system for users, trust providers, and agent providers, generating their respective encryption key pairs; the second stage involves users authenticating themselves through the trust provider, obtaining basic identity credentials, and establishing their unique identifier in the physical world; the third stage involves the agent provider generating a cryptographic commitment for the user's personal agent capability traceability and using a hash chain structure to immutably bind it to the user's identity; the fourth stage involves the trust provider and agent provider collaborating through a security protocol to generate regulatory factors, which are then directly embedded into the identity credentials. Users, holding these credentials, engage in privacy-protected interactions via two-factor DID, and, when necessary, support conditional identity retrospection by regulatory agencies to achieve compliance accountability.

2. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 1, characterized in that, User authentication and identity credential initialization are as follows: User-generated unique anchoring factor and random binding value and encapsulate it as an encrypted credential request. Send to trusted provider TP; After the trust provider verifies the validity of the request, it calculates the intermediate regulatory factor. It also uses a customized chameleon hash algorithm to bind intermediate regulatory factors with anchoring factors; simultaneously, it constructs a distributed identifier for users based on verification information. The trust provider generates a Personality Certificate (PHC) and its corresponding authorization context. and the encrypted response result Return to user; Users decrypt the response results and verify the integrity of the binding relationship between regulatory factors and anchoring factors.

3. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 1, characterized in that, Anchoring factor Request for credentials The calculation is as follows: ； ； Where H represents the hash function, The encryption algorithm indicates the use of a public key. Encrypt; This is a signature algorithm, indicating the use of the user's private key. Sign up. This indicates that the binding factor is randomly generated by the user, BI represents the user's biometric identifier, and PII represents the user's generated personal identity information.

4. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 3, characterized in that, First, the trust provider receives the credential request. Then, using its private key Decryption: ； in, This indicates the use of TP's private key. The algorithm used for decryption; Next, verify the embedded signature. and anchoring factor The correctness; After successful verification, TP uses its Paillier public key. Homomorphic encryption is performed on the identity hash value to calculate the intermediate state regulatory factor. : ； in, A value randomly selected for TP; Then, TP utilizes a custom chameleon hash. The algorithm will Bind to In the process, new random values ​​are generated. : ;in Indicated by A custom chameleon hash algorithm is executed as the trapdoor key; the anchor factor is updated to: ,in Indicates will Binding middle; After binding is complete, TP uses the key derivation function. Calculate the encryption key and utilize Encryption generates secure information packets It was then uploaded to the InterPlanetary File System (IPFS) to obtain a content identifier. ; Calculate the encrypted content identifier using the Paillier encryption algorithm : ； in, This refers to the Paillier encryption algorithm, used in Paillier cryptography. Perform encryption operations on encrypted data.

5. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 3, characterized in that, To issue a user's unique personality credential, the trust provider TP first initializes a personality credential template. This template is a structured instance with all fields initially set to empty, and logically consists of the following four parts. (1) Data owned by the agent itself: a composite field containing identity data maintained by the agent; The composite field is divided into two subfields: TP management issuance item and AP management issuance item; (2) TP-related proof: a field used to record verification evidence issued by the trust provider; (3) AP-related proofs: fields used to record verification evidence issued by the agent provider; (4) Personality Certificate Verification: Security field, used to record all signatures of the personality certificate; Specifically, TP assigns attribute values ​​to the ASO.TPM field of the personality credential. and generate proof ; TP selects two random factors And generate the Chameleon hash value of the trust provider based on the TP public key. Simultaneously, TP uses AP's public key to generate the chameleon hash value for the agent provider. Subsequently, a binding proof is generated by signing the concatenation of the two hash values ​​mentioned above. The TP fills in the relevant information into the corresponding fields of the personality credential; secondly, in order to achieve compliant updates of the PHC by the AP, the TP generates context information for subsequent authorization. : ； TP uses identity hash value As an index, related parameters Store the encrypted final response in the local database. Send to the user.

6. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 5, characterized in that, User received Then, using its private key Decryption: ; Subsequently, the user performs data extraction and verification checks; first, the user extracts anchoring factors from the Personality Certificate (PHC). And extract intermediate regulatory factors from the authorization context TPAC. Finally, users verify the anchoring factor using the following formula. Binding consistency: If all verifications pass, it means that the initial PHC has been successfully generated, and the anchoring of the user's real identity with the PHC has been completed.

7. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 5, characterized in that, The specific steps for verifying the generation of a user's personal intelligent agent are as follows: First, the user sends an encrypted agent request to apply for configuration of a personal agent. Then, the agent provider receives the user's identity credentials and verifies the binding relationship between the user's identity and anchoring factors, constructing and returning a custom module matrix accordingly. Next, the user decrypts the matrix and provides the selected configuration information. The agent provider generates a complete personal agent based on this selection and updates the identity credentials by modifying the hash chain, ensuring the immutability of the agent's technical traceability. Finally, after verifying the consistency of the agent, the user sends a generation request to the trust provider, which ultimately generates the user's distributed identifier and final identity credentials.

8. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 5, characterized in that, The specific steps for verifying the generation of a user's personal intelligent agent are as follows: STEPD1. The user uses the public key provided by the agent provider. For Personality Certificate (PHC) and User Identity Hash Value Intermediate regulatory factors and the trust provider's authorization context Encryption is performed to construct the agent's request. And send it to the agent provider AP; STEPD2. The agent provider AP receives... Then, using its private key Decryption is performed; subsequently, the AP uses the identity hash value. Search the local database to ensure that there is no existing record for the user, in order to confirm the legitimacy of the request and initiate the verification process; AP extracts anchoring factors from personality credential PHC. With proof ; and from Extracting intermediate regulatory factors Updated binding random value And encrypted random factors The AP performs two key verifications:

1. Based on the TP public key, it verifies the TP proof and the signature of the authorization context using a preset verification algorithm; 2. Verify the correctness of the binding factors; After verification, AP constructs a custom module matrix (CMM) as a configuration matrix containing options for each functional module of the agent; Finally, the AP uses the user's public key. For the construction Encrypt and send the encryption result Return to the user; STEPD3. After receiving the custom module matrix encrypted by the AP, the user uses its private key. Decrypt to obtain optional information about the module: ,in This indicates a decryption algorithm used based on the user's private key. Perform decryption on encrypted data; Users exercise the right to define the logical architecture of the intelligent agent, select target modules from a custom module matrix according to actual needs, and generate vectors representing the indices and descriptions of the selected modules: ,in Indicates the first The index of the specific modules selected in each functional category; Finally, the user utilizes the public key provided by the agent provider. For the selected custom module combination Encryption is performed to construct customized information for the personal intelligent agent. and return it to AP: ; STEPD4. The agent provider (AP) uses its private key. The received encrypted combination of custom modules is decrypted to retrieve the module index; then the corresponding code block is located based on the index, and the complete custom module code is reconstructed. AP executes personalized agent generation method This method generates a complete Personal Agent (PA) using a verified Personality Certificate (PHC) and retrieved code modules. As input, the target functional components are reorganized and encapsulated according to a preset logical topology to instantiate a functional autonomous intelligent agent that is cryptographically strongly bound to the user's personality attributes. AP will customize module code With identity hash value Combined, through the hash chain generation algorithm ( Construct configuration commitment ; AP uses Algorithm I to update PHC, anchoring the agent's configuration code to PHC; after updating the personality credential PHC, AP uses a pseudo-random function to generate intermediate variables. ; Subsequently, combined with intermediate regulatory factors The final regulatory factor is calculated using the Paillier homomorphic encryption algorithm. Ultimately, this enables compliant identity traceability for intelligent agents; AP uses identity hash value As an index, a set of key parameters is stored in the local database, including This is to provide a basis for subsequent auditing and maintenance; AP assigns domain name identifiers to users and using the user's public key The updated Personality Certificate (PHC) and regulatory factors Domain name identifier The final response result is constructed by encrypting a personalized smart PA containing randomly bound values. : ； STEPD5. The user receives the final response from the personal intelligent agent. and use its private key Decrypt to obtain the updated binding random value. Regulatory factors Domain name identifier Ultimate proof of personality And personal intelligent agents (PAs): Subsequently, the user extracts the custom module code corresponding to each module from PA. : ; Users utilize the extracted custom module code and identity hash value Recalculate configuration commitment Users retrieve stored information from their Personality Certificate (PHC). and Perform a consistency check; the user performs final verification to ensure the configuration commitments are fulfilled. Correctly bound to anchor factor middle; If all the above verifications pass, the user can utilize... The regulatory factor, domain name identifier, and updated credentials are encrypted, and a distributed identifier generation request is constructed and sent to the trust provider. ; STEPD6. Trust providers use their private keys Received distributed identifier generation request Decrypt to obtain the relevant content: ; TP constructs self-certification identifier and use its public key Calculate the corresponding encrypted self-authentication identifier ; Subsequently, TP executes Algorithm II to generate the final Personality Certificate (PHC) and sends it to the user. If the user successfully receives and stores the PHC, the user obtains a unique dual-anchored identity (PHC).

9. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 8, characterized in that, Algorithm I is an AP-driven PHC update algorithm, and its specific implementation is as follows: Algorithm I input: Module commitment Identity hash Personality certificate to be modified AP Modification Factor and binding factor ; Algorithm I is executed as follows: Agent credential generation: ; ; ; ; Chameleon Collision Hash Search: ; Search for results that meet the following conditions : ; use Algorithm to obtain : ; Update anchoring factor : ; Algorithm I Output: .

10. The method for constructing a dual-anchored identity for a personal AI agent that integrates autonomy and compliance as described in claim 8, characterized in that, Algorithm II is a TP-driven PHC update algorithm, and its specific implementation is as follows: The input to Algorithm II is: distributed identifier. Personality certificate to be modified and TP modification factor ; Algorithm II is executed as follows: TP proof generation: ; ; ; ; Local data query : ; Search for results that meet the following conditions : , ; Algorithm II output: .

Images