Insurers at Lloyd’s of London have launched a product to cover companies for losses caused by artificial intelligence malfunctions.
Damaging payouts
The products will cover against the cost of court claims if a customer or another third party suffers harm as the result of an AI tool.
In the rush to adopt AI, some tools, including customer service bots, have made embarrassing and costly mistakes.
Virgin Money apologised to a customer in January when one of its chatbots reprimanded a customer for using the word ‘virgin.’ Courier service DPD withdrew its customer service bot when it swore at customers. Last year Air Canada was ordered to honour a discount code a customer service chatbot had made up.
Encouraging AI take up
Karthik Ramakrishnan, the chief executive of Armilla, a start-up insurer covering the cost of AI-related court claims, said many companies had been deterred from taking on AI because of fears that tools such as chatbots would break down leaving them exposed to claims.
AI language models are dynamic, meaning they learn over time. Mistakes are made during the learning process.
The cover is available in the case of chatbots giving incorrect information to customers. Compensation will be based on the probability of degradation.
AI systems that are judged to be excessively prone to breakdown will not be covered.
Competent humans still better than AI
Most AI related errors will currently be hidden from public view, only made visible through social media networks whilst companies negotiated compensation with the aggrieved third party.
Well-trained staff and rigorous risk management systems have always been the best insurance against incompetence. A major public payout may be the embarrassment needed to remind businesses of the limitations of AI.
