New tool balances accuracy with fairness in social media


By Dr. Tim Sandle
June 12, 2025

DIGITAL JOURNAL

Australian legislation could force social media firms to take steps to prevent those under 16 years of age from accessing platforms such as X, TikTok, Facebook and Instagram - Copyright GETTY IMAGES NORTH AMERICA/AFP/File Michael M. Santiago

Earlier this year, Facebook rolled back rules against some hate speech and abuse. Along with changes at X (formerly Twitter) that followed its purchase by Elon Musk, the shifts make it harder for social media users to avoid encountering toxic speech.

That does not mean all social networks and other online spaces have given up on the massive challenge of moderating content to protect users. One novel approach relies on artificial intelligence. AI screening tools can analyze content on large scales while sparing human screeners the trauma of constant exposure to toxic speech.

Yet AI content moderation faces a challenge, according to Maria De-Arteaga, assistant professor of information, risk, and operations management at Texas McCombs. This is with: being fair as well as being accurate.

An algorithm may be accurate at detecting toxic speech overall, but it may not detect it equally well across all groups of people and all social contexts.

“If I just look at overall performance, I may say, oh, this model is performing really well, even though it may always be giving me the wrong answer for a small group,” De-Arteaga explains. For example, it might better detect speech that’s offensive to one ethnic group than to another.

In new research, De-Arteaga and her co-authors show it’s possible to achieve high levels of both accuracy and fairness. What’s more, they devise an algorithm that helps stakeholders balance both, finding desirable combinations of accuracy and fairness for their particular situations. De-Arteaga worked with datasets of social media posts already rated “toxic” and “nontoxic” or safe by previous researchers. The sets totaled 114,000 posts.

The researchers used a fairness measurement called Group Accuracy Parity (GAP), along with formulas that helped train a machine learning model to balance fairness with accuracy. Applying their approach through AI to analyze the datasets:It performed up to 1.5% better than the next-best approaches for treating all groups fairly.
It performed the best at maximizing both fairness and accuracy at the same time.

But GAP is not a one-size-fits-all solution for fairness, De-Arteaga notes. Different measures of fairness may be relevant for different stakeholders. The kinds of data needed to train the systems depends partly on the specific groups and contexts for which they’re being applied.

For example, different groups may have different opinions on what speech is toxic. In addition, standards on toxic speech can evolve over time.

Getting such nuances wrong could wrongly remove someone from a social space by mislabeling nontoxic speech as toxic. At the other extreme, missteps could expose more people to hateful speech.

The challenge is compounded for platforms like Facebook and X, which have global presences and serve wide spectrums of users.

“How do you incorporate fairness considerations in the design of the data and the algorithm in a way that is not just centered on what is relevant in the U.S.?” De-Arteaga says.

For that reason, the algorithms may require continual updating, and designers may need to adapt them to the circumstances and kinds of content they’re moderating, she says. To facilitate that, the researchers have made GAP’s code publicly available.

High levels of both fairness and accuracy are achievable, De-Arteaga says, if designers pay attention to both technical and cultural contexts.

“You need to care, and you need to have knowledge that is interdisciplinary,” she says. “You really need to take those considerations into account.”

The article“Finding Pareto Trade-Offs in Fair and Accurate Detection of Toxic Speech” is published in Information Research.

Searches for DeepSeek increase amid ChatGPT crash


By Dr. Tim Sandle
June 12, 2025

DIGITAL JOURNAL


OpenAI's ChatGPt and DeepSeek are among growing ranks of rivals as tech firms compete to lead in the hot field of generative artificial intelligence models - Copyright AFP Lionel BONAVENTURE

Open AI’s ChatGPT faced a global outage on Tuesday, 10th June 2025, with users from all over the world complaining about error messages. the cause of the crash remains unknown. According to an update on the OpenAI website, the situation is currently under investigation, citing that users are having “elevated error rates and latency” on services, Chat GPT, APIs, and Sora.


ChatGPT is a generative artificial intelligence chatbot launched in 2022 by OpenAI. As opposed to predictive AI, generative AI is trained on large amounts of data in order to identify patterns and create content of its own, including voices, music, pictures, and videos. ChatGPT allows users to interact with the chatting tool much like they could with another human, with the chatbot generating conversational responses to questions or prompts.

Across social media and Slack threads, creative teams admitted to delays. Copy drafts could not be finished and brand decks could not be polished.

This was not bad news for everyone. A Google Trends analysis by experts at QR Code Generator has revealed that searches for other generative artificial intelligence (AI) chatbots have significantly increased amid this ChatGPT blackout.

According to the findings, searches for DeepSeek are projected to reach 2.13 million queries today (10th June), a 109% increase compared to past the 30-day daily average of 1.02 million.

DeepSeek refers to a new set of frontier AI models from a Chinese startup of the same name. DeepSeek has caused quite a stir in the AI world this week by demonstrating capabilities competitive with – or in some cases, better than – the latest models from OpenAI.

Searches for Claude AI are also expected to increase significantly, rising from 149,441 average daily searches in the past 30 days to 291,181 today, a 95% increase.

The experts also reveal that searches for Google Gemini and Microsoft Copilot are projected to increase by 80% and 52%, respectively.

Marc Porcar, CEO of QR Code Generator PRO S.L, has told Digital Journal: “While ChatGPT has captured significant market share, many competitors have been steadily building their user bases and capabilities.

“Today’s global outage of ChatGPT shows the importance of diversification and the challenges that come when the world relies on one single platform and it experiences a blackout.

“The immediate 109% surge in DeepSeek searches and 95% increase for Claude shows that users are actively seeking alternatives rather than simply waiting. This ultimately benefits the AI ecosystem by redistributing market attention and proving that no single platform, regardless of popularity, should be considered indispensable or the sole source for critical workflows.”