IBM is going for trust - again!
An internal project was recently uncovered again and formulated in a focused way: Firstly, IBM employees want to be seen as knowledgeable - about their customers' industries and required solutions; secondly, they want to be seen as being able to build solutions based on continuous innovation; and finally they want to be a trusted partner. Trust should be the glue that holds the company together internally and also with customers, partners, the ecosystem, and society.
In another area of trust, IBM is trying to convince customers to trust IBM because of its security offerings. IBM’s IT security efforts are massive. They did not seriously engage in these activities until 2015, but IT security has become a substantial business since (about $2bn in 2017). The open framework built around its security platform invites partners to contribute to more secure IT environments. An IT security war room and the X-Force report round off the offerings.
Moreover, IBM cares about customer data more than any other cloud company. Their commitment to not touching customer data and not using it for internal purposes or analysis clearly sets IBM apart from its cloud competitors. By promising that customer data belongs to customers, the company has another substantial asset for differentiating its offering. In the past, IBM surprised the industry with the Open Cloud Manifesto (2014), and now with its commitment to the EU Commission’s Cloud COC (Code of Conduct). It clearly says: Do not touch customer data. It would be great if the other major cloud vendors adopted a similar approach.
Last but not least, IBM is on its way to making AI comprehensible, explainable, and a tool for a trusted partnership between man and machine instead of keeping AI a black box where nobody understands why a certain recommendation is given. An example: people want to know why their credit scoring was not good enough to get a mortgage. It will not be enough for the banker to simply tell the potential customer that “the algorithm” told us. Especially in the age of GDPR, this can be a major differentiator. AI has to be transparent in the way results are produced.
Finally, AI is only as good as the training data. For humans, about 180 aspects have been identified in which decisions may be biased. In order to ensure that AI is independent from human – i.e. programmer/trainer – bias, training bodies have to be unbiased and neutralized. Under GDPR rules, they also have to be anonymized. This will lead to a loss of the information body. However, the dividend is higher trust, which IBM thinks is the underpinning of our dealing with AI and business in general.
Mentioning details about the mainframe Hardware Security Modules (HSM) here would go too deep into the technology. However, the inherent inclusion of HSMs into the CPU (of the mainframe computers) makes it extremely difficult to compromise data stored on these systems. Newer solutions also address a combination of cloud deployments, VMs and the usage of HSMs seamlessly.
Bottom line: IBM is digging out a basic law of good business practices: trust! It is the 3000-year-old formula that delivers the best long-term relationships.