Since its launch in November 2022, the AI chatbot ChatGPT (chat generative pre-trained transformer), which was built on top of the GPT-3 large language model, has been trending online and in print.
It has been described as “even more massive than you think” in a podcast by Josh Bersin and “a window into the real future of financial services”, according to a Forbes article written by David Birch. But it attracted more mainstream headlines on 17 January, when musician Nick Cave, after being inundated by fans with lyrics created by the bot in his own writing style, expressed his views on one of them in a blog post. “This song sucks,” was his trenchant verdict, according to an article in Dazed magazine, written by Thom Waite.
Tech giant backing
Cave may have been underwhelmed by ChatGPT’s creative efforts, but others have high hopes for its transformative potential. On 23 January, Microsoft announced a “new multi-year, multibillion-dollar investment” with Open AI, the bot’s creator, according to an article on CNBC by Ashley Capoot. “The deal marks the third phase of the partnership between the two companies, following Microsoft’s previous investments in 2019 and 2021,” writes Capoot. Microsoft CEO Satya Nadella is quoted in the article, saying that the partnership was formed to advance research in AI and democratise the technology.
Microsoft has not yet revealed details of how the bot could be integrated into its existing products, but there has been speculation as to potential use cases. It could be used to act as a personal assistant, generate correspondence and respond to no-code programming suggestions within the Office 365 suite of applications, according to Gartner vice president and analyst Bern Elliot, quoted in an article on Reworked by David Barry.
Risk and reward
The financial services sector, too, is assessing the potential of the technology. In 2020, the percentage of mid-size banks and credit unions deploying chatbots more than tripled from 4% to 13% , according to an article in Forbes by Ron Shevlin. Now, the authors of a new book on risk modelling and AI, aimed at financial risk practitioners, claim that AI and machine learning can “help banks and insurers take on their most vexing challenges”, according to a PR Newswire report. The authors aim to “demystify AI and [machine learning] through practical guidance and real-world examples”.
The article quotes Terisa Roberts, one of the book’s co-authors, as saying: “Adopting AI and machine learning to mitigate credit, market, liquidity and other emerging risks is paramount.” However, in the book’s introduction, the authors sound a note of caution, advising that for the technology to “realise short- and long-term business value in a responsible way, the foundational technological building blocks of data, people, and processes will need to be reconsidered”.
Who’s responsible?
Their caution would appear to be justified. A new report from the global analytics software platform provider FICO, conducted by Corinium and based on a survey of 100 banking and financial AI executives, concludes that, although demand for the technology is surging, financial services firms “lack responsible AI strategies”. Reporting on the findings, Businesswire highlights four key takeaways:
- 52% of financial services organisations see AI as more important than a year ago
- Only 8% of respondents have a ‘fully mature’ scaleable AI model
- 43% struggle to meet regulatory requirements
- 71% do not have ethical and responsible AI as a core part of their operational strategies.
The findings, the report concludes, suggest that AI has the potential to mitigate risk and build stronger relationships with customers, and that enhanced levels of customer trust could in turn boost market share – but it emphasises that using the technology responsibly is key to deriving these benefits.