Is your tech language racist?

Our language often reflects how we view the world and the habits we have formed. It's essential in our everyday lives and at work. But what if our language negatively impacts another person?

Language in technology and cybersecurity has been developed over time between teams as short hand to describe a specific process, activity or operation. The language is often technical and can be used to document how a system works. There are many terms in technology and cyber security that have strong negative associations and using them creates barriers based on ethnicity, gender, sexual orientation, and accessibility.

Which is why we're working with UK Finance and industry peers Microsoft and EY to address this. A new paper, Use of non-inclusive language in technology and cybersecurity and why it matters, highlights the industry’s commitment to driving change and ensuring language is inclusive and promotes diversity in the workplace.
 

Where non-inclusive language comes from

Language is contextual and evolves over time. Being mindful of how a word has become corrupted through misuse in context is a good first step in objectively identifying a broad suite of potential terms that may require replacing.

As language evolves, its use evolves, and it is important to reflect on its current contextual use to promote diverse and inclusive working practices and be mindful about the effect that language may have on people. Terms now deemed unacceptable were previously acceptable, and it is conceivable that as cultural norms and social values continue to evolve, that other terms may become unacceptable over time.

For so long the tech space has used outdated and inherently problematic words and phrases. Our language reflects who we are and who we want to be, so for Lloyds Banking Group to be a truly inclusive organisation, we needed to change. While change doesn’t happen overnight, the Group have been at the forefront of this work at an industry level and a tangible difference can be seen in how we communicate. 

Why does language matter?

Language affects how people see the issues they are trying to solve and can impact decision-making.

As language builds assumptions, these then need to be broken down and challenged when trying to manage technology and cybersecurity differently. Bias in language and visual cues have also been identified as key ethical considerations in the training phases of machine learning and facial recognition, to establish pattern recognition, and false positives. The potential for reinforced biases is an important factor to consider when designing and building technologies, in order to acknowledge and remove any potential ‘poisoning’ of algorithms.

Some language is steeped in negativity that highlights and entrenches inequality, further compounding a system that often creates barriers based on race and ethnicity, gender, sexual orientation, and accessibility. It is incumbent on all of us to ensure that we remove those barriers, creating a society that is fairer and equal to all.

At Lloyds Banking Group, we continue to invest in an environment where everyone can bring their whole selves to work and experience psychological safety. Setting this ethos has helped set standards of acceptance which enrich how we interact and speak with our customers from all walks of life.  

Making our tech language inclusive

In many situations, education is the first step in going about change. Below are some terms which are outdated, along with some suggested alternatives.

When proposing alternatives in the moment, it’s extremely powerful to do so in an encouraging way where the other person understands the context of why, whilst feeling more supported, than embarrassed, to learn a new phrase. This supports the second principle of respect (see below) and inspires change instead of a collision of opinions. 

Example outdated terms and suggested alternatives

3 principles of respect

Paul Vincent, Head of IT Cyber Security, has suggested three values to help people take specific action when it comes to inclusive tech language:
 

• Think before you speak. Your words count. How will they be perceived by those you are speaking to or those observing? Avoid phrases that link ‘Black’ with negative situations (e.g. Blacklists, Blackout, Black Sheep, Black Swan). Lead by example but be cautious about creating rules for others. Be curious about common phrases and share what you learn with others in the spirit of understanding. 
• Do not be quick to assume malice. If someone says something you feel offended by, do not be quick to assume they intended it to be hurtful. Feel empowered to call out offensive terms/phrases, but do so in a kind, considerate and empathic way. Confide in someone you have a good relationship with if you do not feel you can speak to the person yourself. 
• Be quick to apologise. It doesn’t matter about who was in the right or wrong, be quick to apologise if you have upset someone by something that was said. Make sure you are sincere and learn from the experience. We all say things that are taken the wrong way or blurted out in a clumsy manner; it’s how we deal with it afterwards that defines us.

Microsoft, who have contributed to the paper, share how they're making their tech language more inclusive:

"Microsoft is committed to designing products and services that are both inclusive and accessible to all that use them. But this isn’t just about the end result, inclusion needs to be imbedded at the very core. That’s why we ensure that our tech language is also inclusive. For example, we have worked with stakeholders to change our technical documents, changing references to master/slave terminology, replacing the word ‘slave’ to ‘secondary’ or other appropriate terms. Longer-term, we’re partnering with others to use the technology industry’s influence as a driving force to deliver positive, systematic change." Monica Rush, Senior Program Manager, Microsoft

Going forward

This is the beginning of this journey for us. Working with our UK Finance partners, we must continue to raise awareness of inclusive tech language and drive forward change across the industry.

Changing the process and the policy was the easy bit but the biggest hurdle has been getting people to change the terms they use of out habit. We know this will take time, but people are now more aware and equipped better to challenge these terms. As we continue to remove non-inclusive terms from our documents and process, change will happen.

It’s clear from the research and discussions that there is much more that can be done to challenge our inherited biases. Getting creative and recognising when we get it wrong ourselves, even in front of others, is a great first step to leading by example, especially to enrich both corporate and social communities. How will we each make a difference; what alternative terms can you come up with?