Artificial intelligence Joins the Fight Against Global Corruption

Given that Artificial intelligence (AI) has been useful in so many ventures, anti-corruption scholars are eager to apply it to their work. In fact, AI has been described as “the next frontier in anti-corruption.” AI holds a number of benefits for anti-corruption work and is particularly good at detecting corrupt activities that were formerly undetectable by human efforts — all at a much faster rate.

According to an article originally published on The Conversation, an independent and nonprofit source of news, analysis and commentary from academic experts, an integrated corporate social responsibility (CSR) and artificial intelligence (AI) approach can help mitigate corruption risks.

For example, the Tax Administration Services of Mexico used AI algorithms and analytical tools to detect 1,200 fraudulent businesses and 3,500 fraudulent transactions within a three-month period. This would have taken about 18 months to achieve without AI assistance. However, AI and anti-corruption discussions so far have mostly focused on governmental efforts to address corporate corruption, not on companies using AI to mitigate corporate corruption — even though many of them already use AI to maximize profit. Using CSR as an anti-corruption tool involves reinforcing the need for companies to fill governance gaps in anti-corruption regulation and not exploit the lack of adequate legal and institutional frameworks when operating in the Global South, the authors maintain.

Adopting this approach means companies have to go beyond the bare minimum, by not just complying with anti-corruption legislation, but viewing anti-corruption as a social responsibility. Doing so will center transparency and accountability at the heart of a company’s operations and communicate to stakeholders and shareholders that the company places a premium on these values. In the corporate anti-corruption context, AI can provide companies with a proposed investment destinations or transactions and help detect corruption risks in such ventures and improve due diligence processes.
AI can also provide more information for yearly anti-corruption policy reviews and assist in designing training based on AI analyses of company processes, reports and operations.

However, there are ethical concerns that arise from AI use. These concerns include violations of the right to privacy and freedom of expression, AI data biases or the replication of bias in AI. There are also concerns that such technologies in themselves can be used as tools to facilitate corruption as corrupt persons may use AI to fine-tune their processes.

A Brazilian machine learning application called MARA is a case study in such concerns. This application involves evaluating the probability of an individual acting corruptly based on data retrieved from their social security number. This raises privacy concerns, considering social security numbers contain sensitive information. The application also depends on criminal conviction data, which may be biased since marginalized groups are often over-represented in the criminal justice system. An application like MARA is likely to proliferate this bias.

READ MORE ARTICLES