Trustworthy AI
   HOME

TheInfoList



OR:

Trustworthy AI is a programme of work of the
ITU The International Telecommunication Union is a specialized agency of the United Nations responsible for many matters related to information and communication technologies. It was established on 17 May 1865 as the International Telegraph Unio ...
(United Nations Specialized Agency for ICT) under its
AI for Good AI for Good is a year-round digital platform of the United Nations, where AI innovators and problem owners learn, discuss and connect to identify practical AI solutions to advance the UN SDGs. The impetus for organizing global summits that are act ...
programme. The programme advances the standardization of a number of
Privacy-enhancing technologies Privacy-enhancing technologies (PET) are technologies that embody fundamental data protection principles by minimizing personal data use, maximizing data security, and empowering individuals. PETs allow User (computing), online users to protect the ...
(PETs), including
homomorphic encryption Homomorphic encryption is a form of encryption that permits users to perform computations on its encrypted data without first decrypting it. These resulting computations are left in an encrypted form which, when decrypted, result in an identical ...
,
federated learning Federated learning (also known as collaborative learning) is a machine learning technique that trains an algorithm across multiple decentralized edge devices or servers holding local data samples, without exchanging them. This approach stands in ...
,
secure multi-party computation Secure multi-party computation (also known as secure computation, multi-party computation (MPC) or privacy-preserving computation) is a subfield of cryptography with the goal of creating methods for parties to jointly compute a function over their ...
,
differential privacy Differential privacy (DP) is a system for publicly sharing information about a dataset by describing the patterns of groups within the dataset while withholding information about individuals in the dataset. The idea behind differential privacy is t ...
,
zero-knowledge proof In cryptography, a zero-knowledge proof or zero-knowledge protocol is a method by which one party (the prover) can prove to another party (the verifier) that a given statement is true while the prover avoids conveying any additional information a ...
. Privacy-Enhancing Technologies are complex and the working can be considered counterintuitive. For instance, homomorphic encryption enciphers data, yet if this data is numerical, it is still possible to compute on this data (the outcomes still being encrypted but decryptable by the data owner). Other PETs exhibit similar characteristics. Often these technologies are developed with the explicit goal of enabling their analysis in juristictions different from the data creation (under e.g.
GDPR The General Data Protection Regulation (GDPR) is a European Union regulation on data protection and privacy in the EU and the European Economic Area (EEA). The GDPR is an important component of EU privacy law and of human rights law, in parti ...
). As such, the programme being led by two
international organizations An international organization or international organisation (see spelling differences), also known as an intergovernmental organization or an international institution, is a stable set of norms and rules meant to govern the behavior of states a ...
develops
international standards international standard is a technical standard developed by one or more international standards organizations. International standards are available for consideration and use worldwide. The most prominent such organization is the International Org ...
to work in this context. The PETs are used from analytics such as
Artificial Intelligence Artificial intelligence (AI) is intelligence—perceiving, synthesizing, and inferring information—demonstrated by machines, as opposed to intelligence displayed by animals and humans. Example tasks in which this is done include speech r ...
.


History

The origin of the programme lies with the
ITU-WHO Focus Group on Artificial Intelligence for Health The ITU-WHO Focus Group on Artificial Intelligence for Health (AI for Health) is an inter-agency collaboration between the World Health Organization and the ITU, which created a benchmarking framework to assess the accuracy of AI in health. Th ...
, where strong need for privacy at the same time as the need for analytics, created a demand for a standard in these technologies. When
AI for Good AI for Good is a year-round digital platform of the United Nations, where AI innovators and problem owners learn, discuss and connect to identify practical AI solutions to advance the UN SDGs. The impetus for organizing global summits that are act ...
moved online in 2020, the TrustworthyAI seminar series was initiated to start discussions on such work, which eventually led to the standardization activities.


Standardization


Multi-Party Computation

MPC is being standardizated under "Question 5" (the incubator) of
ITU-T Study Group 17 The ITU-T Study Group 17 (SG17) is a statutory group of the ITU Telecommunication Standardization Sector (ITU-T) concerned with security. The group is concerned with a broad range of security-related standardization issues such as cybersecuri ...
.


Homomorphic Encryption

ITU has been collaborating since the early stage of the HomomorphicEncryption.org standardization meetings, which has developed a standard on
Homomorphic encryption Homomorphic encryption is a form of encryption that permits users to perform computations on its encrypted data without first decrypting it. These resulting computations are left in an encrypted form which, when decrypted, result in an identical ...
. The 5th homomorphic encryption meeting was hosted at ITU HQ in
Geneva , neighboring_municipalities= Carouge, Chêne-Bougeries, Cologny, Lancy, Grand-Saconnex, Pregny-Chambésy, Vernier, Veyrier , website = https://www.geneve.ch/ Geneva ( ; french: Genève ) frp, Genèva ; german: link=no, Genf ; it, Ginevr ...
.


Federated Learning

Zero-sum masks as used by federated learning for privacy preservation are used extensively in the multimedia standards of
ITU-T Study Group 16 The ITU-T Study Group 16 (SG16) is a statutory group of the ITU Telecommunication Standardization Sector (ITU-T) concerned with multimedia coding, systems and applications, such as video coding standards. It is responsible for standardization of ...
(
VCEG The Video Coding Experts Group or Visual Coding Experts Group (VCEG, also known as Question 6) is a working group of the ITU Telecommunication Standardization Sector (ITU-T) concerned with standards for compression coding of video, images, audio ...
) such as JPEG,
MP3 MP3 (formally MPEG-1 Audio Layer III or MPEG-2 Audio Layer III) is a coding format for digital audio developed largely by the Fraunhofer Society in Germany, with support from other digital scientists in the United States and elsewhere. Origin ...
, and
H.264 Advanced Video Coding (AVC), also referred to as H.264 or MPEG-4 Part 10, is a video compression standard based on block-oriented, motion-compensated coding. It is by far the most commonly used format for the recording, compression, and distr ...
,
H.265 H is the eighth letter of the Latin alphabet. H may also refer to: Musical symbols * H number, Harry Halbreich reference mechanism for music by Honegger and Martinů * H, B (musical note) * H, B major People * H. (noble) (died after 12 ...
(aka
MPEG The Moving Picture Experts Group (MPEG) is an alliance of working groups established jointly by ISO and IEC that sets standards for media coding, including compression coding of audio, video, graphics, and genomic data; and transmission and f ...
).


Zero-knowledge proof

Previous pre-standardization work on the topic of zero-knowledge proof has been conducted in the ITU-T Focus Group on Digital Ledger Technologies.


Differential privacy

The application of differential privacy in the preservation of privacy was examined at several of the "Day 0" machine learning workshops at AI for Good Global Summits.


See also

*
Artificial Intelligence Artificial intelligence (AI) is intelligence—perceiving, synthesizing, and inferring information—demonstrated by machines, as opposed to intelligence displayed by animals and humans. Example tasks in which this is done include speech r ...
*
Privacy-enhancing technologies Privacy-enhancing technologies (PET) are technologies that embody fundamental data protection principles by minimizing personal data use, maximizing data security, and empowering individuals. PETs allow User (computing), online users to protect the ...
* Data Science


References

{{reflist Information privacy United Nations Economic and Social Council International Telecommunication Union