Old Twitter Vs X: Israel-Gaza War Spotlights 'Information Crisis'

Users of X, formerly Twitter, complain they can no longer decipher truth, as captured by news photographers, from fiction on the site. MOHAMMED ABED / AFP/File
Users of X, formerly Twitter, complain they can no longer decipher truth, as captured by news photographers, from fiction on the site. MOHAMMED ABED / AFP/File
TT
20

Old Twitter Vs X: Israel-Gaza War Spotlights 'Information Crisis'

Users of X, formerly Twitter, complain they can no longer decipher truth, as captured by news photographers, from fiction on the site. MOHAMMED ABED / AFP/File
Users of X, formerly Twitter, complain they can no longer decipher truth, as captured by news photographers, from fiction on the site. MOHAMMED ABED / AFP/File

Twitter won fame in the Arab uprisings nearly a decade ago as a pivotal source for real-time crisis information, but that reputation has withered after the platform's transformation into a magnet for hate speech and disinformation under Elon Musk.

Historically, Twitter's greatest strength was as a tool for gathering and disseminating life-saving information and coordinating emergency relief during times of crisis. Its old-school verification system meant sources and news were widely trusted, said AFP.

Now the platform, renamed X by new owner Musk, has gutted content moderation, restored accounts of previously banned extremists, and allowed users simply to purchase account verification, helping them profit from viral -- but often inaccurate -- posts.

The fast-evolving Israel-Gaza conflict has been widely seen as the first real test of Musk's version of the platform during a major crisis. For many experts, the results confirm their worst fears: that changes have made it a challenge to discern truth from fiction.

"It is sobering, though not surprising, to see Musk's reckless decisions exacerbate the information crisis on Twitter surrounding the already tragic Israel-Hamas conflict," Nora Benavidez, senior counsel at the watchdog Free Press, told AFP.

The platform is flooded with violent videos and images -- some real but many fake and mislabeled from entirely different years and places.

Nearly three-fourths of the most viral posts promoting falsehoods about the conflict are being pushed by accounts with verified checkmarks, according to a new study by the watchdog NewsGuard.

In the absence of guardrails, that has made it "very difficult for the public to separate fact from fiction," while escalating "tension and division," Benavidez added.

'Fire hose of information'
That was evident on Tuesday after a deadly strike on a hospital in war-ravaged Gaza, as ordinary users scrambling for real-time information vented frustration that the site had become unusable.

Confusion reigned as fake accounts with verified checkmarks shared images of past conflicts while peddling hasty conclusions of unverified videos, illustrating how the platform had handed the megaphone to paying subscribers, irrespective of accuracy.

Accounts masquerading as official sources or news media stoked passions with inflammatory content.

Misinformation researchers warned that many users were treating an account of an activist group called "Israel war room," stamped with a gold checkmark –- indicating "an official organization account," according to X –- as a supposedly official Israeli source.

India-based bot accounts known for anti-Muslim rhetoric further muddied the waters by pushing false anti-Palestinian narratives, researchers said.

Meanwhile, Al Jazeera warned that it had "no ties" to an account that falsely claimed affiliation to the Middle East broadcaster as it urged its followers to "exercise caution."

"It has become incredibly challenging to navigate the fire hose of information -- there is a relentless news cycle, push for clicks, and amplification of noise," Michelle Ciulla Lipkin, head of the National Association for Media Literacy Education, told AFP.

"Now it's clear Musk sees X not as a reliable information source but just another of his business ventures."

The chaos stands in sharp contrast to the 2011 Arab uprisings that prompted a surge of optimism in the Middle East about the potential of the platform to spread authentic information, mobilize communities and elevate democratic ideals.

'Break the glass'
The breakdown of the site's basic functionality threatens to impede or disrupt the humanitarian response, experts warn.

Humanitarian organizations have typically relied on such platforms to assess needs, prepare logistical plans and assess whether an area was safe to dispatch first responders. And human rights researchers use social media data to conduct investigations into possible war crimes, said Alessandro Accorsi, a senior analyst at the Crisis Group.

"The flood of misinformation and the limitations that X put in place for access to their API," which allow third-party developers to gather the social platform's data, had complicated those efforts, Accorsi told AFP.

X did not respond to AFP's request for comment.

The company's chief executive Linda Yaccarino has signaled that the platform was still serious about trust and safety, insisting that users were free to adjust their account settings to enable real-time sharing of information.

But researchers voiced pessimism, saying the site has abandoned efforts to elevate top news sources. Instead, a new ad revenue sharing program with content creators incentivizes extreme content designed to boost engagement, critics say.

Pat de Brun, head of Big Tech Accountability at Amnesty International said X should use every tool available, including deploying so-called "break the glass measures" aimed at dampening the spread of falsehoods and hate-speech.

"Platforms have clear responsibilities under international human rights standards," he told AFP.

"These responsibilities are heightened in times of crisis and conflict."



China’s Zhipu AI Launches Free AI Agent, Intensifying Domestic Tech Race

Words reading "Artificial intelligence AI", miniature of robot and toy hand are pictured in this illustration taken December 14, 2023. (Reuters)
Words reading "Artificial intelligence AI", miniature of robot and toy hand are pictured in this illustration taken December 14, 2023. (Reuters)
TT
20

China’s Zhipu AI Launches Free AI Agent, Intensifying Domestic Tech Race

Words reading "Artificial intelligence AI", miniature of robot and toy hand are pictured in this illustration taken December 14, 2023. (Reuters)
Words reading "Artificial intelligence AI", miniature of robot and toy hand are pictured in this illustration taken December 14, 2023. (Reuters)

Chinese artificial intelligence startup Zhipu AI unveiled a free AI agent on Monday, joining a wave of similar launches in China's increasingly competitive AI market.

The product, called AutoGLM Rumination, can perform deep research as well as tasks including web searches, travel planning, and research report writing, CEO Zhang Peng said at a lunch event in Beijing.

The agent is powered by Zhipu's proprietary models, including its reasoning model GLM-Z1-Air and foundation model GLM-4-Air-0414. The company claims GLM-Z1-Air matches rival DeepSeek's R1 in performance while running up to eight times faster and requiring only one-thirtieth of the computing resources.

AI agents are systems designed to make decisions and execute a range of tasks autonomously.

The launch follows a surge in Chinese AI product releases after DeepSeek shook the industry earlier this year with a model that it said operated at substantially lower costs than US rivals.

It also comes weeks after competitor Manus sparked interest with what it marketed as the world's first general AI agent.

While Manus charges users up to $199 monthly, Zhipu's AutoGLM Rumination will be available free of charge through the company's official channels, including its GLM model website and mobile app.

Zhipu AI, founded in 2019 as a spinoff from a Tsinghua University laboratory, has emerged as one of China's leading AI startups.

The company, which developed the GLM series of models, claims its latest large language model GLM4 outperforms OpenAI's GPT-4 on several benchmarks.

The startup made headlines earlier this month after securing three consecutive rounds of government-backed funding in a single month. The most recent investment came from the city of Chengdu, which injected 300 million yuan ($41.5 million) into the company.