Former Google X CBO urges for more high-quality Arabic content to train AI

Gawdat also recognized the opportunities that AI brings for authors and researchers. (AFP/File)
Short Url
Updated 29 May 2023
Follow

Former Google X CBO urges for more high-quality Arabic content to train AI

  • Mo Gawdat expressed concerns about scarcity of rich online content and AI ability to provide valuable insights into the Arab world

LONDON: Artificial intelligence programs like ChatGPT should be trained with more high-quality Arabic content, Egyptian writer and entrepreneur Mo Gawdat said.

Speaking during the International Congress for Arabic Publishing and Creative Industries event in Abu Dhabi, Gawdat expressed concern about the scarcity of rich Arabic discourse online.

The author of “Solve for Happy: Engineer Your Path to Joy” and former chief business officer of Google X, the company’s experimental arm, argued that without deeper Arabic content, AI applications may struggle to provide valuable insights into the Arab world.

“A lot of megabytes of Arabic language online are very empty,” Gawdat explained. “The initial versions of AI will reflect whatever it is that we put out there.”

Even though he believes that future AI iterations will become smart enough to discern between “real information and fluff,” he said the lack of quality content should be tackled as soon as possible.

Gawdat called on Arabic content creators, publishers, and educational institutions to produce more insightful material for AI tools to analyze.

“The problem we are facing is that we are creating an impression that knowledge of the Arab world is shallow, in comparison to the incredible wisdom that is really coming from the region,” he said.

“We need to create content with depth and with more intellectual value.”

Gawdat added that as future AI systems increasingly rely on authoritative sources, educational institutions should make more Arabic research accessible online.

“With AI utilizing the Arabic content that we put in, it is our responsibility to put the right kind of content out there,” he emphasized.

“This will ultimately determine the quality of Arabic artificial intelligence-driven content.”

Gawdat compared this responsibility to parenting, describing the process as being aware of how children learn to interact with the world.

He also urged researchers and content creators to reconsider their strategies, as the ChatGPT-led revolution in information access and content creation has the potential to threaten human authors.

However, Gawdat also recognized the opportunities that AI brings.

“As an author, I have learned to no longer define myself as someone who is supposed to write words on paper,” he said.

“My job definition now is to provide interesting human perspectives and inspiration regarding certain topics.

“Through whatever knowledge is available to me and in whatever form, I am trying to find a way to connect with other humans and communicate what I believe needs to be communicated.”


Study finds nearly half of UK news stories on Muslims show signs of bias

Updated 09 March 2026
Follow

Study finds nearly half of UK news stories on Muslims show signs of bias

  • Centre for Media Monitoring finds 20,000 out of 40,913 articles from 30 major news outlets contain bias and 70% link Muslims to negative behaviors or themes
  • Findings reveal ‘deeply concerning evidence of structural bias’ in portrayal of Muslims by UK press and point to ‘systemic problem’ within the media, says center’s director

LONDON: Nearly half of news articles published in the UK in 2025 that referenced Muslims or Islam contained some degree of bias, according to a report issued on Monday by the Centre for Media Monitoring. It also found that about 70 percent of stories linked Muslims to negative behaviors or themes.

The nonprofit organization, which tracks the ways in which Muslims and Islam are portrayed in the media, examined 40,913 articles from 30 major news outlets and found that about 20,000 showed some form of bias.

The study looked at “structural patterns” in coverage that “shape public narratives” about Muslims amid rising hostility toward the community.

“As the largest study of its kind ever conducted in the UK, this report presents deeply concerning evidence of structural bias in how Muslims are portrayed in the UK press,” said Rizwana Hamid, the director of the organization.

It found that 70 percent of the articles it reviewed highlighted negative aspects related to Muslims, though not all of the stories were biased in themselves. The wider patterns were also troubling: 44 percent of the coverage omitted key context, 17 percent relied on generalizations, and 13 percent included outright misrepresentation.

Taken together, the monitoring center said, the findings amounted to evidence of an “information integrity crisis” that distorts public understanding, and “a deeply concerning trend” in reporting on Muslims.

The research points to a “systemic problem within our media ecosystem,” Hamid said.

“When entire communities are repeatedly framed through lenses of suspicion or threat, it inevitably shapes public attitudes, political debate and the everyday lives of British Muslims,” she added.

News brands targeting right-wing audiences were more likely to produce biased coverage, the report found.

The Spectator magazine and GB News were identified as having the highest proportion of “very biased” articles, and as the “worst across all five bias categories”: negative framing, generalizations, misrepresentation, lack of context, and problematic headlines.

Other outlets highlighted for displaying high levels of biased content about Muslims included The Telegraph, The Jewish Chronicle, Daily Express, The Sun, Daily Mail and The Times.

In contrast, the BBC, other broadcasters and left-leaning outlets recorded the lowest rates of bias in the study.

The research comes as British Muslims report rising levels of discrimination. Official figures published in October revealed that religious hate crimes against Muslims rose by 19 percent in the year to March 2025 compared with the previous 12 months.