In recent days, , Aopeng China was awarded by the international authoritative certification body BSI Issued by the ISO 9001 and ISO 27701 certificate , Covering aopeng （ China ） Shanghai headquarters 、 Wuxi and Dalian delivery center . This is the information security management certification obtained by aopeng China before ISO 27001 after , Once again, it has achieved international recognition in terms of quality control and data privacy protection " Gold standard ".

The British Standards Institution

BSI It is the world's leading management system audit 、 Professional certification training and product quality inspection institutions , To be objective 、 fair 、 The principle of justice is global 193 A country's 90,000 Many locations offer professional 、 Independent certification and other services .

ISO 9001 Quality management standards

ISO 9001 It is one of the most mature quality control systems in the world , In the world 161 More than 75 It is widely used in 10000 enterprises and organizations .

ISO 27001 Information security management system

ISO 27001 It is the international standard of information security management . It defines how to implement an independently evaluated and certified information security management system , It can effectively ensure the security of financial and confidential data , And reduce the possibility of unauthorized access .

ISO 27701 Privacy authentication

ISO 27701 It is a high standard international management system , It aims to provide guidance for enterprises and organizations to protect personal privacy , Fulfill the high commitment to personal information management and protection .

It's not just a privacy leak

What is high-quality data ？ If this question was answered a few years ago , I believe most of the answers will revolve around “ accuracy ” and “ Efficient ”. However , Today, when big data is already pervasive , More people are beginning to “ Security ”、“ responsible ” Keywords regarded as high-quality data .

A few years ago 3·15 A face recognition information disclosure event exposed at the party , It is still fresh in people's memory . In recent years , More and more enterprises and consumers begin to think about the risks of personal information being abused or even infringing on the rights and interests of others . However , responsible AI It's not just the protection of privacy , It also includes being unbiased 、 Safety compliance 、 Interpretable and positive social impact . stay AI During the deployment of the project , The risks that enterprises need to consider mainly include ：

User experience and business risks

for example , A credit card company wants to develop a AI Program , Used to help determine whether to issue a credit card to a customer . Some discrimination in the process of building the model , It will hurt some user groups , Thus, it brings great losses to the business .

Brand reputation risk

A report from MIT shows , At present, some commercial facial recognition technologies on the market , Most white men have a high recognition rate ; For women with dark skin, the recognition accuracy is very low . This is bound to have a negative impact on brand reputation .

Legal risk

When AI The model is constantly commercialized , Will face different laws and compliance requirements from different countries and regions . The data management team must have knowledge of both security and privacy issues , Also fully understand the local laws and regulations . otherwise , Violating relevant local laws and regulations AI Application may face serious consequences .

Big guy said so

Aopeng Appen Through the network of industry experts , Ask them about AI government 、 Privacy issues and “ responsible AI” And other relevant industry hot topics ：

Decades ago , In smartphones 、 Before the Internet came to the fore , Data has only one home ： database . Now data is almost everywhere , Data privacy and security must continue to evolve , Adapt to new computing needs , Data can be managed 、 Both transfer and analysis can be protected .

——Protegrity Data and AI Executive Vice President Eliano Marques

More and more enterprises are worried AI Group bias in Algorithm （ race 、 Age 、 Gender, etc ） The impact on the brand and the possibility of legal problems arising therefrom . When evaluating data suppliers , High quality and unbiased data should be the primary screening condition .

——Jumio CEO Robert Prigge

One ‘ responsible AI’ The new era of . responsible AI Set standards and best practices for responsible training data , Aimed at Improve the quality 、 Efficiency and transparency , Including eliminating bias in training data , While promoting inclusiveness and collaboration . In turn, , This will also increase AI The value to the business , And the public for the use of AI The trust of the .

—— Aopeng Appen Chief technical officer Wilson Pang

responsible AI Training data

Aopeng Appen Think , With AI Explosive growth of the industry , responsible AI The whole AI Ecosystem considerations , Become AI Methods actively adopted in all links of the industrial chain . in fact , Built in a responsible way AI Model , It is often more successful in practical applications .

The problem of bias needs to be vigilant at the beginning of model construction , Otherwise, it may be introduced into all stages of development and production . In the data annotation phase , Both the data itself and the data annotator need to realize diversification . Aopeng's global 100 More than ten thousand skilled crowdsourcing resources come from 170+ Countries and regions , Have more than 292 Professional knowledge of languages and dialects . meanwhile , Aopeng Appen One stop shop for AI Auxiliary data annotation platform MatrixGo, Adopt a bias minimization framework , In the way of combining machine and man , Provide real-time and on-demand intelligent crowdsourcing annotation process .

Click here to , View aopeng Appen Crowdsourcing ethics .

For data quality control 、 Security management and privacy protection , Aopeng Appen Always committed to providing customers with the highest level of management standards . Except as mentioned at the beginning ISO 9001、ISO 27001 and ISO 27701 Beyond certification , Worldwide , Aopeng Appen Also passed GDPR,SOC 2 Type II,HIPAA And other data security compliance certification in different countries and regions around the world , Ensure that data sources and channels are formal 、 Security 、 legal . Besides , Aopeng Appen Annotation platform MatrixGo It also has authority control 、 Data encrypted transmission 、PII Strict data security control strategies such as information encryption and storage . The platform scans vulnerabilities every day , Maximize the security of customer data assets .

Click here to , View the detailed aopeng Appen Privacy statement .

Big data era , How to provide intelligent humanized services and maintain personal privacy 、 Find a balance between information security , It is a topic worthy of discussion from all walks of life . In our country , Data resources have become national basic strategic resources and innovative elements of social production . Following “ Network security law ” after , A series on data security 、 The introduction of laws and regulations on Information Protection , It will further accelerate the construction of data quality control and security assurance system in the intelligent era .

In this regard , Aopeng Appen Will continue to practice the high standards of the industry , With high-quality data and maturity 、 Sustainable management processes jointly usher in an era of responsible intelligence .