Food for thought: Is OpenAI exploiting uploaded images?

 

Dr Evans Sagomba
Everything AI

 

OVER the past few months, people across the world, including Zimbabwe, have been caught up in a viral trend: uploading personal photos to platforms such as ChatGPT to receive unique AI-generated avatars.

 

At face value, this activity seems like harmless fun, a chance to see your image transformed into a whimsical artistic design.

 

However, beneath the surface lies a troubling reality: OpenAI is using these voluntarily uploaded images to enhance its AI algorithms, bypassing crucial privacy concerns and exploiting users’ ignorance.

The implications of this trend go far beyond the creation of avatars.

 

OpenAI’s privacy policy allows the company to collect and use personal data submitted by users to train its AI models unless users actively opt out.

 

For Zimbabweans, this raises serious questions.

 

How are these images being used?

 

What happens to them after they are uploaded?

 

And most importantly, what privacy risks are involved?

How OpenAI collects and uses images

When users upload their images to platforms like ChatGPT, they often assume they are engaging in a simple transaction, a photo for a stylised avatar.

 

But OpenAI is not just generating fun artistic designs.

 

The images users upload is stored and integrated into vast datasets used to train and refine AI algorithms.

 

These algorithms rely on high-quality data to improve accuracy, continuity, and realism in image generation, and OpenAI has found a clever way to collect such data for free.

OpenAI’s privacy policy explicitly states that it uses user-submitted data to enhance its models unless individuals opt out of this process.

 

This means that when you upload your photo, you are likely contributing to the AI’s broader capabilities.

 

What is worse, many Zimbabweans may not even be aware that this is happening, let alone how to opt out.

Scraped images versus voluntarily uploaded photos

You might ask, “What’s the harm? Can’t OpenAI simply scrape the same images from the internet?”

 

This is not entirely true, for two reasons.

 

First, in places like the European Union (EU), stringent data protection laws such as the General Data Protection Regulation (GDPR) create legal barriers to scraping personal images from the internet.

OpenAI can only process such data if it can prove a legitimate interest — a lawful basis that ensures the rights and privacy of individuals are respected.

 

In practice, these regulations mean OpenAI often has to refrain from using scraped images or take additional protective measures when dealing with sensitive data, such as images of minors.

However, when users voluntarily upload their photos, a different legal ground applies: consent.

 

By agreeing to the platform’s terms and conditions, individuals grant OpenAI the freedom to use their data without the need for extensive privacy safeguards.

 

In Zimbabwe, where no robust data privacy laws currently exist, this voluntary submission bypasses even the minimal protections that might otherwise apply.

Secondly, voluntarily uploaded images provide OpenAI with something far more valuable than scraped images: freshness and originality.

 

These photos often include new content, such as family pictures or intimate moments, which have not yet appeared on social media.

 

In other words, OpenAI is gaining access to exclusive images, enriching its datasets with high-quality, diverse visual material that other platforms don’t have.

How did OpenAI make this trend so popular?

It wasn’t an accident.

 

OpenAI leveraged a carefully orchestrated marketing strategy to ensure the viral success of this trend.

 

Four days before its announcement of ChatGPT’s upgraded image generator, OpenAI CEO Sam Altman shared a stylised, AI-generated Ghibli-style image on X (formerly Twitter) with the provocative quote, “Feel the AGI.”

 

Altman’s post triggered immediate interest, with his 3.5 million followers sharing the image widely and creating their versions soon after.

This PR manoeuvre ignited what is now being referred to as the “Ghibli Effect.” Millions of people began uploading personal photos to ChatGPT to generate similar artistic avatars, with many voluntarily submitting images that had never appeared online before.

 

While users celebrated their creative outputs, OpenAI quietly gained access to valuable original images without spending a penny.

Why this matters for Zimbabwe?

For Zimbabweans, this trend is especially concerning due to the country’s lack of strong data protection laws.

 

Unlike in regions such as the EU, there are no robust regulations in Zimbabwe to oversee how companies like OpenAI collect, store, and use personal data.

 

This leaves users vulnerable to exploitation and misuse, with little recourse to challenge the practices of global tech giants.

The implications of this exploitation are significant.

 

First, privacy risks are heightened as sensitive images—such as family photos—are stored indefinitely, with no guarantees of secure handling.

Secondly, Zimbabweans are unwittingly contributing to the growth of a billion-dollar industry without any compensation or control over their personal data.

 

Finally, the absence of regulatory frameworks means there are no protections in place to prevent the misuse of these images for profiling, manipulation, or other harmful activities.

What can Zimbabweans do to protect themselves?

Until Zimbabwe develops comprehensive data privacy laws, individuals must take steps to safeguard their personal information.

 

Here are some practical measures:

Read the Fine Print: Familiarise yourself with the terms and conditions of platforms like ChatGPT to understand how your data may be used.

 

Opt Out: OpenAI provides an option to opt out of having your data used for training its AI models.

 

Take advantage of this feature to limit how your images are handled.

Avoid uploading sensitive photos

Refrain from submitting family pictures, intimate photos, or images of minors to any platform unless necessary.

Spread Awareness: Educate friends and family about the risks of uploading personal images, encouraging informed decisions online.

The need for government action

The government of Zimbabwe must urgently prioritise the creation and implementation of data privacy regulations that protect citizens’ rights.

 

These regulations should include:

Transparent consent mechanisms to ensure users are fully aware of how their data is being collected and used.

 

Limits on data usage, preventing companies from using personal photos for purposes beyond those initially intended.

Accountability measures requiring companies to disclose how and where user data is stored.

A wake-up call for Zimbabweans

Participating in viral trends may seem harmless, but it’s essential to look beyond the surface and understand the broader implications.

 

When you upload your images to platforms like ChatGPT, you are giving away more than a photograph—you are surrendering control over your data, feeding into a system that thrives on exploitation and global inequality.

For Zimbabweans, the absence of robust data privacy laws makes this issue even more urgent.

 

Until protections are implemented, citizens must take proactive steps to safeguard their digital identities.

Remember, that stylised avatar might come at a cost you’re not prepared to pay the price of your privacy.

 

Let this be a call to action for individuals, communities, and policymakers alike to recognise the value of our data and defend our rights in the digital age.

A call for action: Zimbabwe needs urgent AI governance to safeguard its future

Time waits for no one, and for Zimbabwe, the clock is ticking fast when it comes to regulating Artificial Intelligence (AI).

 

The longer the nation delays implementing clear AI governance frameworks, the more deeply embedded the challenges of AI misuse will become.

 

As AI tools evolve at breakneck speed, their potential for misuse, such as forgery, privacy violations, and social manipulation, is expanding just as rapidly.

 

This is a clarion call: Zimbabwe must act now to avert an impending crisis.

For the past three months, I have been actively advocating for the Government of Zimbabwe (GOZ), through the Ministry of Information Communication Technology, Postal and Courier Services (MICTPCS), to prioritise the development and implementation of a Zimbabwe Artificial Intelligence Governance Regulatory Framework (ZAIGRF).

Such a framework is no longer a luxury; it is a necessity for building trust, aligning AI technologies with Zimbabwe’s societal values, and ensuring that AI serves the collective interests of its citizens rather than becoming a tool for harm.

To oversee the effective implementation of ZAIGRF, I have further proposed the establishment of an independent and specialised authority, the Zimbabwe Artificial Intelligence Regulatory Authority (ZAIRA).

ZAIRA would serve as the country’s guiding hand in navigating the challenges and opportunities posed by AI.

 

Its role would involve not only enforcing ethical guidelines but also promoting safe innovation that delivers meaningful benefits to all Zimbabweans.

If you have more questions, send them to the editors or direct them to me, and I will respond to them in next week’s issue.

Dr Evans Sagomba, MPhil, PhD, is a Doctor of Philosophy, who specialises in AI, Ethics and Policy Researcher, AI Governance and Policy Consultant, Ethics of War and Peace Research Consultant, Political Philosophy, and also a Chartered Marketer/CMktr, FCIM, [email protected], Social media handles; LinkedIn; @ Dr Evans Sagomba (MSc Marketing)(FCIM )(MPhil) (PhD) /X: @esagomba.