AI: The risks and opportunities for the title free mission

Artificial intelligence (AI) is now embedded in a huge variety of aspects of modern life, from healthcare and hiring systems to customer service and social media platforms.

As these systems become more sophisticated, the way they handle personal identity information—especially gender identity and marital status titles should be a key focus in EDI conversations.

AI has the power to either support inclusive practices… or reinforce outdated social norms.

When they aren’t carefully designed, AI systems risk marginalising colleagues and customer based on gender assumptions and marital stereotypes.

How does AI impact gender identity divesity and our marital status title cause?

…And what can be done to address it?

🔍 The Problem: Binary and Traditional Defaults

Very limited Gender Options in Data Fields
- Most AI systems are trained using binary gender categories: male/female.
- Nonbinary, transgender, intersex, and gender-fluid individuals are often completely forgotten, or forced to select inaccurate labels.
Constant use of Gendered Titles
- Many forms and digital assistants still default to titles like Mr., Mrs., Miss, which are gendered. They imply marital status only for women, meaning women are not being being treated equitably when titles are demanded.
- These titles are often mandatory selections, with no gender-neutral alternatives like “Mx.” available… and no options to choose nothing, to bypass the box or to free type a title which better represents non binary people.
Inflexible Marital Status Options
- Systems frequently ask for marital status with traditional categories: single, married, divorced, widowed.
- These options do not account for:
  - Civil unions
  - Domestic partnerships
  - Cohabitation without legal marriage
  - Relationship choices that don’t conform to heteronormative norms

⚠️ The Risks and Consequences

Misgendering and Misclassification
- AI systems that make assumptions based on names or voice can misgender users.
- This leads to incredible emotional harm, alienation, and mistrust of digital platforms.
Exclusion from Services
- Users who do not see themselves represented in drop-down menus or chatbot responses may abandon applications, surveys, or customer service chats. (71% of people surveyed by GoTitleFree said they would be put off buying a product or service if they were addressed incorrectly).
- IF a person abandons that process, they will be limiting their own access to healthcare, financial services, and employment.
Amplification of Bias
- If an AI system is trained on historical data that reflects past discrimination or gender imbalance, it will replicate that bias.
- Example: CV screening AI preferring “David” over “Davina” or ignoring resumes with non traditional names or gender markers.

💡 Opportunities for Inclusion

AI systems don’t have to perpetuate these issues. There is a huge opportunity here!

With thoughtful design, they can become tools for inclusion and equity.

✅ 1. Expanding Gender and Title Options

Include a wide variety of gender identities:
- Options such as nonbinary, genderqueer, agender, transgender, or prefer not to say.
- Allow open-text fields for users to self-describe their gender. Because sometimes people want to say, but they are restricted by the options in the list and don’t have an option to self describe.
Offer inclusive titles:
- Allow users to opt out of using titles altogether if unnecessary. Or at the very least, add “Mx.” as a gender-neutral option.

✅ 2. Removing Unnecessary Gender and Marital Data Collection

Only collect gender or marital status if it’s essential for the service. (Ask yourself WHY it is necessary for the service. WHY do you really need to know?)
Avoid assuming that such data is required for identity verification or personalisation. Remember that titles are not a part of our legal identity. They are purely social and not stated on a person’s passport.

✅ 3. Customizable Identity Display

Let users set their preferred name, pronouns, and title in profiles and communications.
Reflect chosen identity across all AI interactions (email, chatbots, automated replies, etc.).

✅ 4. Bias Detection and Mitigation

Regularly audit algorithms for signs of gender or relationship bias.
Use diverse training data that includes a broad range of identities and relationship models.
Implement human-in-the-loop systems for sensitive tasks, allowing real-time corrections.

🌍 Real-World Applications and Examples

➤ Social Media Platforms

Platforms like LinkedIn now allow users to display pronouns and choose from more gender options.
Facebook introduced over 50 gender identity options to better represent user diversity.

➤ HR and Recruitment Tools

AI hiring platforms are being redesigned to eliminate gendered language from job ads.
Some systems now ignore marital status and gender markers to reduce discrimination.

➤ Healthcare Systems

Electronic health records in some places are being updated to reflect gender identity separately from biological sex.
Inclusive AI in healthcare can help to improve care for transgender and nonbinary patients, but progress has been slow to non-existent, and the trans community needs more advocacy than ever before.

➤ Government and Legal Forms

Some governments have finally introduced gender-neutral ID options (e.g., “X” instead of “M” or “F”).
Public platforms are beginning to allow gender-neutral and relationship-neutral identifiers. But it’s been slow progress, and as we know, governments take us backwards as often as they take us forwards.

📢 The Role of Developers and Policymakers

👨‍💻 For Developers:

Collaborate with LGBTQ+ advocacy groups when you are designing new systems and platforms
Include gender-diverse testers in the quality assurance process.
Design with flexibility and user self-definition in mind.

🏛️ For Policymakers:

Mandate inclusive data standards in public-sector platforms.
Support transparency and explainability in AI decision-making. (i.e. ensure that any decisions based on AI generated sources are understood, and that sources are clear!)
Create legal protections against algorithmic discrimination based on gender identity or marital status.

🔮 The Future of Inclusive AI

AI has the potential to be a powerful ally in the fight for equity—if it’s trained and implemented thoughtfully. Systems that are inclusive of all gender identities and relationship models will:

Build trust with users
Promote dignity and self-expression
Reduce structural barriers for marginalised communities

There is a great opportunity ahead, and by challenging traditional assumptions around gender and marriage, AI can contribute to a society that celebrates individuality rather than erasing it.

✅ Key Takeaways

AI sadly often defaults to binary gender roles and traditional marital status labels, which can exclude or misrepresent people.
Truly inclusive AI requires:
- More gender options
- Neutral titles… or preferably no titles at all!
- Voluntary data sharing
- Bias audits and diverse training data
Real-world systems are already adapting, but wider systemic change is really needed if we are to adapt well, and mindfully.
Both developers and regulators play incredibly important roles in ensuring AI respects and reflects the full and wonderful diversity of human identity.

By redesigning systems with these principles in mind, AI can move from being a huge barrier to inclusion, to becoming a powerful tool for progression.

__________________________________________________________________

GoTitleFree™ has been celebrating businesses who have gone ‘title free’ for seven years.

We have created toolkits for businesses to use who want to explore the various ways to ensure gender inclusive language in customer service and communication practices.

We also have a petition for any person who wishes to add their name to our list of supporters.

To all of our existing supporters, keep fighting the ‘title free’ fight!

Stella Sutcliffe x