Bias in artificial intelligence development has been a growing concern as its use increases across the world. But despite efforts to create\u00a0AI standards, it is ultimately down to organizations and IT leaders to adopt best practices and ensure fairness throughout the AI life cycle to avoid any dire regulatory, reputation, and revenue impact, according to a new Forrester Research\u00a0report.\n\nWhile a 100% elimination of bias in AI is impossible, CIOs must determine when and where AI should be used and what could be the ramifications of its usage, said Forrester vice\u00a0president\u00a0Brandon Purcell.\n\nBias has become so inherent in AI models that companies are looking at bringing in a new C-level executive called the chief ethics officer tasked with navigating the ethical implications of AI, Purcell said. Salesforce, Airbnb, and Fidelity already have ethics officers and more are expected to follow suit, he told CIO.com.\n\nEnsuring AI model fairness\n\nCIOs can take several steps to not only to measure but also balance AI models\u2019 fairness, he said, even though there is a lack of regulatory guidelines dictating the specifics of fairness.\n\nThe first step, Purcell said, is make sure that the model itself is fair. He recommended using accuracy-based fairness criterion[GG3] that optimizes for equality, a representation-based fairness criterion that optimizes for equity, and an individual-based fairness criterion. Companies should bring together multiple fairness criteria to check the impact on the model\u2019s predictions.\n\nWhile the accuracy-based fairness criterion ensures that no group in the data set receives preferential treatment, the equity-based fairness criterion ensures that the model is offering equitable results based on the data sets.\n\n\u201cDemographic parity, for example, aims to ensure that equal proportions of different groups are selected by an algorithm. For example, a hiring algorithm optimized for demographic parity would hire a proportion of male to female candidates that is representative of the overall population (likely 50:50 in this case), regardless of potential differences in qualifications,\u201d Purcell said.\n\nOne example of bias in AI was the\u00a0Apple Card AI model\u00a0that\u00a0was allocating more credit to men, as was revealed in late 2019. The issue came to light when the model offered Apple cofounder Steve Wozniak a credit limit that was 10 times than that of his wife even though they share the same assets.\n\nBalancing fairness in AI\n\nBalancing the fairness in AI across its life cycle is important to ensure that a model\u2019s prediction is close to being free of bias.\n\nTo do so, companies should look at soliciting feedback from stakeholders to define business requirements, seek more representative training data during data understanding, use more inclusive labels during data preparation, experiment with causal inference and adversarial AI in the modeling phase, and accounting for intersectionality in the evaluation phase, Purcell said. \u201cIntersectionality\u201d refers to how various elements of a person\u2019s identity combine to compound the impacts of bias or privilege.\n\n\u201cSpurious correlations account for most harmful bias,\u201d he said. \u201cTo overcome this problem, some companies are starting to apply causal inference techniques, which identify cause-and-effect relationships between variables and therefore eliminate discriminatory correlations.\u201d Other companies are experimenting with adversarial learning, a machine-learning technique that optimizes for two cost functions that are adversarial.\n\nFor example, Purcell said, \u201cIn training its VisualAI platform for retail checkout, computer vision vendor Everseen used adversarial learning to both optimize for theft detection and discourage the model from making predictions based on sensitive attributes, such as race and gender. In evaluating the fairness of AI systems, focusing solely on one classification such as gender may obscure bias that is occurring at a more granular level for people who belong to two or more historically disenfranchised populations, such as non-white women.\u201d\n\nHe gave the example of Joy Buolamwini and Timnit Gebru\u2019s seminal paper on algorithmic bias in facial recognition that found that the error rate for Face++\u2019s gender classification system was 0.7% for men and 21.3% for women across all races, and that the error rate jumped to 34.5% for dark-skinned women.\n\nMore ways to adjust fairness in AI\n\nThere are couple of other methods that companies might employ to ensure fairness in AI that include deploying different models for different groups in the deployment phase and crowdsourcing with bias bounties \u2014 where users who detect biases get rewarded \u2014 in the monitoring phase.\n\n\u201cSometimes it is impossible to acquire sufficient training data on underrepresented groups. No matter what, the model will be dominated by the tyranny of the majority. Other times, systemic bias is so entrenched in the data that no amount of data wizardry will root it out. In these cases, it may be necessary to separate groups into different data sets and create separate models for each group,\u201d Purcell said.