Quick Buzz Feed
IT service provider is using AI to increase throughput while applying it to ‘big ticket’ challenges externally.
CGI started with ChatGPT in a relatively small way 18 months ago, with 250 licenses. The company is now rolling out enterprise ChatGPT access to its entire 9,000-strong UK workforce, from operations and project managers to finance, commercial, and HR teams, as well as software developers. Russell Goodenough, UK head of AI at CGI, emphasizes that everyone can use the tool to increase speed and throughput. CGI has a long history with machine learning and data science, but ChatGPT's emergence in 2023 led to the creation of a dedicated AI centre of excellence. The company monitors AI use in its operations and observes significant productivity improvements, with automated testing achieving a five-times boost and rapid prototyping environments showing even greater gains. These improvements are being leveraged to grow the business and address 'big ticket' problems in critical national infrastructure rather than to cut jobs.
To ensure successful AI adoption, Goodenough states that CGI has intentionally avoided a top-down or even centre-down approach. Instead, he collaborates with a network of 'evangelists and enthusiasts' across the business, proactively communicating the positives of AI and highlighting internal successes. This strategy has proven effective, particularly among coders, who have reported substantial time savings and a five-fold increase in productivity within the past year. While the UK region leads initiatives with OpenAI, other regions within CGI's global 92,000-strong workforce spearhead efforts with other AI platforms, such as Google Gemini in APAC and Anthropic Claude in Canada. The insights and results from these regional leads are shared across all regions, ensuring widespread access to AI advancements. Currently, around 18,000 CGI employees use ChatGPT, with 8,000 based in the UK.
Despite the global investment and rapid adoption of AI, Goodenough expresses concern over the direction some AI development is taking. He questions why resources are being allocated to generate inappropriate content, such as images of women without clothes, especially when there are significant societal challenges to address. Goodenough emphasizes his commitment to applying AI for beneficial purposes, particularly to tackle major issues in areas like policing, justice, and healthcare. He believes that the technology should be used to 'do good' and solve critical national problems rather than being misused for trivial or harmful applications.