Skip to content Skip to footer
Search

Loading Results

PwC's Responsible AI

AI you can trust

AI is bringing limitless potential to push us forward as a society — but with great potential comes great risks. 

When you use AI to support business-critical decisions based on sensitive data, you need to be sure that you understand what AI is doing, and why. Is it making accurate, bias-aware decisions? Is it violating anyone’s privacy? Can you govern and monitor this powerful technology? Globally, organisations recognise the need for Responsible AI but are at different stages of the journey.

Responsible AI (RAI) is the only way to mitigate AI risks. Now is the time to evaluate your existing practices or create new ones to responsibly and ethically build technology and use data, and be prepared for future regulation. Future payoffs will give early adopters an edge that competitors may never be able to overtake.

PwC’s RAI diagnostic survey can help you evaluate your organisation’s performance relative to your industry peers. The survey takes 5-10 minutes to complete and will generate a score to rank your organisation with actions to consider.

Take our free Responsible AI Diagnostic

Responsible AI is the leading priority among industry leaders for AI applications in 2021, with emphasis on improving privacy, explainability, bias detection, and governance. 

Among respondents to our AI Predictions survey, three quarters of companies have formal or some ethical guidance or policies in place, and 1 in 5 have an ethical framework in place. 35% of respondents have plans to improve governance of AI systems and processes in 2021. 

22% of US companies said they are ‘definitely’ able to deploy AI at scale compared to only 8% of UK companies. This difference indicates there are opportunities to address limitations in model operations and ability to govern systems.

Source: PwC US - 2021 AI Predictions & 2021 Responsible AI Insights Report

Potential AI Risks

A variety of factors can impact AI risks, changing over time, stakeholders, sectors, use cases, and technology. Below are the six major risk categories for application of AI technology. 

Performance

AI algorithms that ingest real-world data and preferences as inputs may run a risk of learning and imitating possible biases and prejudices.

Performance risks include:

  • Risk of errors
  • Risk of bias and discrimination
  • Risk of opaqueness and lack of interpretability
  • Risk of performance instability

Security

For as long as automated systems have existed, humans have tried to circumvent them. This is no different with AI.

Security risks include:

  • Adversarial attacks 
  • Cyber intrusion and privacy risks
  • Open source software risks

Control

Similar to any other technology, AI should have organisation-wide oversight with clearly-identified risks and controls.

Control risks include:

  • Lack of human agency
  • Detecting rogue AI and unintended consequences
  • Lack of clear accountability

Economic

The widespread adoption of automation across all areas of the economy may impact jobs and shift demand to different skills.

Economic risks include:

  • Risk of job displacement
  • Enhancing inequality
  • Risk of power concentration within one or a few companies

Societal

The widespread adoption of complex and autonomous AI systems could result in “echo-chambers” developing between machines, and can have broader impacts on human-human interaction.

Societal risks include:

  • Risk of misinformation and manipulation
  • Risk of an intelligence divide
  • Risk of surveillance and warfare

Enterprise

AI solutions are designed with specific objectives in mind which may compete with overarching organisational and societal values within which they operate. Communities often have long informally agreed to a core set of values for society to operate against. There is a movement to identify sets of values and thereby the ethics to help drive AI systems, but there remains disagreement about what those ethics may mean in practice and how they should be governed. Thus, the above risk categories are also inherently ethical risks as well. 

  • Enterprise risks include:
  • Risk to reputation
  • Risk to financial performance
  • Legal and compliance risks
  • Risk of discrimination
  • Risk of values misalignment

PwC’s Responsible AI Toolkit

Your stakeholders, including board members, customers, and regulators, will have many questions about your organisation's use of AI and data, from how it’s developed to how it’s governed. You not only need to be ready to provide the answers, you must also demonstrate ongoing governance and regulatory compliance.

Our Responsible AI Toolkit is a suite of customizable frameworks, tools and processes designed to help you harness the power of AI in an ethical and responsible manner - from strategy through execution. With the Responsible AI toolkit, we’ll tailor our solutions to address your organisation’s unique business requirements and AI maturity.

Our Responsible AI Toolkit addresses the three dimensions of Responsible AI

Who is accountable for your AI system? The foundation for Responsible AI is an end-to-end enterprise governance framework, focusing on the risks and controls along your organization’s AI journey—from top to bottom. PwC developed robust governance models that can be tailored to your organisation. The framework enables oversight with clear roles and responsibilities, articulated requirements across three lines of defense, and mechanisms for traceability and ongoing assessment.

Are you anticipating future compliance? Complying with current data protection and privacy regulation and industry standards is just the beginning. We monitor the changing regulatory landscape and identify new compliance needs your organization should be aware of, and support the change management needed to create tailored organizational policies and prepare for future compliance.

How are you identifying risk? You need expansive risk detection and mitigation practices to assess development and deployment at every step of the journey, and address existing and newly identified risks and harms. PwC’s approach to Responsible AI works with existing risk management structures in your organization to identify new capabilities, and supports the development of any necessary operating models. 

 

Is your AI unbiased? Is it fair? An AI system that is exposed to inherent biases of a particular data source is at risk of making decisions that could lead to unfair outcomes for a particular individual or group. Fairness is a social construct with many different and—at times—conflicting definitions. Responsible AI helps your organisation to become more aware of bias and potential bias, and take corrective action to help systems improve in their decision-making.

 

How was that decision made? An AI system that human users are unable to understand can lead to a “black box” effect, where organisations are limited in their ability to explain and defend business-critical decisions. Our Responsible AI approach can help. We provide services and processes to help you explain both overall decision-making and also individual choices and predictions, and we can tailor to the perspectives of different stakeholders based on their needs and uses.

How will your AI system protect and manage privacy? With PwC’s Responsible AI toolkit, you can identify strategies to lead with privacy considerations and to respond to consumers’ evolving expectations. 

What are the security risks and implications that should be managed? Detecting and mitigating system vulnerabilities is critical to maintaining integrity of algorithms and underlying data while preventing the possibility of malicious attacks. The great possibilities of AI come with the need for great protection and risk management. PwC’s approach to Responsible AI includes essential cybersecurity assessments to help you manage effectively. 

Will your AI behave as intended? An AI system that does not demonstrate stability, and consistently meets performance requirements, is at increased risk of producing errors and making the wrong decisions. To help make your systems more robust, Responsible AI includes services to help you identify potential weaknesses in models and monitor long-term performance. PwC has developed specific technical tools to support this area.

Is your AI safe for society? AI system safety should be evaluated in terms of potential impact to users, ability to generate reliable and trustworthy outputs, and ability to prevent unintended or harmful actions. PwC’s Responsible AI services enable you to assess safety and societal impact to support this dimension. 

Is your data use and AI ethical? Our Ethical data and AI Framework provides guidance and a practical approach to help your organisation with the development and governance of AI and data solutions that are ethical and moral. 

As part of this dimension, our framework includes a unique approach to contextualising and applying ethical principles, while identifying and addressing key ethical risks.

Are you positioning your AI toward future compliance? As the regulatory landscape continues to evolve, maintaining compliance and responding to regulatory change will be critical. Leveraging PwC’s approach to Responsible AI can help you identify and evaluate relevant policy, industry standards and regulations that may impact your AI solutions. Operationalize regulatory compliance while factoring in localized differences.

Trust is core to our purpose at PwC.   

We have the right team to build AI responsibly internally and for our clients, and bring big ideas to life across all stages of AI adoption.

  • AI, Data & Data Use Governance
  • AI and Data Ethics
  • AI Expertise: Machine Learning, Model Operations & Data Science
  • Privacy
  • Cybersecurity
  • Risk Management 
  • Change Management
  • Compliance and legal
  • Sustainability and Climate Change
  • Diversity and Inclusion

PwC is proud to collaborate with the World Economic Forum to develop forward-thinking and practical guidelines for AI development in use across industries. Learn more here.

We support all phases of the RAI journey

Assess: Technical and qualitative assessments of models and processes to identify gaps
 
Build: Development and design of new models and processes, given a specific need and opportunity
 
Validate + Scale: Technical model validation and deployment services; governance and ethics change management
 
Evaluate + Monitor: Readiness for AI including confirming controls framework design, internal audit training

Innovate responsibly

Whether you're just getting started or are getting ready to scale, Responsible AI can help. Drawing on our proven capability in AI innovation and deep global business expertise, we'll assess your end-to-end needs, and design a solution to help you address your unique risks and challenges.

Contact us

Contact us today. Learn more about how to become an industry leader in the responsible use of AI.

Recognition & Awards

  • 2020 World Changing Idea, Responsible AI Toolkit, FastCompany
  • Ranked Leader for AI Consulting, Forrester
  • Outstanding achievement in Enterprise, Adoption of AI and AI Ethics, CogX
  • Ranked Leader for Data & Analytics Services, Gartner
  • 100 Brilliant Women in AI Ethics

Contact us

Anand Rao

Anand Rao

Global Artificial Intelligence Leader, Principal, PwC United States

Tel: +1 (617) 633 8354

Euan Cameron

Euan Cameron

UK Artificial Intelligence and Drones Leader, PwC United Kingdom

Tel: +44 (0)7802 438423

Prof. Matt Kuperholz

Prof. Matt Kuperholz

Partner, Chief Data Scientist, PwC Australia

Tel: +61 (3) 8603 1274

Annie Veillet

Annie Veillet

Partner, One Analytics, PwC Canada

Tel: +1 514 205 5146

Hide