Both Microsoft and Google are committed to the use of artificial intelligence (AI) to improve how their applications work. However, these tech companies are also acutely aware that bad AI could potentially damage their brands.
How AI is used
Microsoft has added numerous AI features to Office 365 and SharePoint. One example of how this is used is classification suggestions for filing shared documents. Google uses AI to refine searches to display results that are more relevant.
Form 10-K disclosures
Form 10-K documents are forms that US law requires companies to file. They are an overall view of a business and its finances. One segment of the documents covers risk factors, and Microsoft and Google in their 10-K filings have stated that AI has risks as well as benefits.
Alphabet, the parent company of Google, said in its 10-K form that new products and services which use AI present both technological and legal challenges that could negatively affect their brand and therefore impact badly on their revenue.
In its own 10-K form, Microsoft admitted that AI algorithms can be flawed, can contain biased information and involve controversial data practices, which could be sufficient to impair the acceptance of AI solutions. Microsoft says that flawed AI could lead to legal action against the company or harm their brand. If AI solutions are seen to infringe human rights, privacy or social issues, AI could, in effect, turn people against Microsoft.
By admitting that AI could be harmful, these tech companies are warning investors about the possible impact bad AI can have on their profits. They are being open to their shareholders about the possible dangers of AI.
Why now?
Google and Microsoft have been developing AI for a number of years. Google’s self-driving cars use AI, and Microsoft’s Cortina assistant in Windows 10 was developed with AI features. This raises the question of why the two tech companies are admitting to the dangers of AI now and not previously.
Years ago, Microsoft developed a chatbot called Tay, which started making racist comments, and is still held up as an example of AI going wrong. Though there have been isolated incidents in the past like this example, it is only over recent years that the public has started to become aware of the negative side of AI.
The Cambridge Analytica’s use of Facebook data and AI to influence voters in the 2016 Presidential Election created a scandal that caused many people to realise the dangers of the wrong use of AI.
Dealing with the danger
Google and Microsoft’s admittance that AI can be used in harmful ways should cause them to focus on the issue. The alternative could be that governments regulate the use of AI, although tech companies currently prefer voluntary controls to legally imposed ones.
At WM Reply, we recognise that AI in SharePoint and Office 365 can be a great benefit to business production and worker collaboration, as long as it is done in a responsible way and can help businesses stay ahead with their use of technology. If you would like to find out more about how AI-enhanced software could help your business, get in touch with our team at WM Reply today.