Elon Musk's Grok shows 'flaws' in fact-checking Israel-Iran war: study
Elon Musk's AI chatbot Grok produced inaccurate and contradictory responses when users sought to fact-check the Israel-Iran conflict, a study said Tuesday, raising fresh doubts about its reliability as a debunking tool.
With tech platforms reducing their reliance on human fact-checkers, users are increasingly utilizing AI-powered chatbots -- including xAI's Grok -- in search of reliable information, but their responses are often themselves prone to misinformation.
"The investigation into Grok's performance during the first days of the Israel-Iran conflict exposes significant flaws and limitations in the AI chatbot's ability to provide accurate, reliable, and consistent information during times of crisis," said the study from the Digital Forensic Research Lab (DFRLab) of the Atlantic Council, an American think tank."Grok demonstrated that it struggles with verifying already-confirmed facts, analyzing fake visuals, and avoiding unsubstantiated claims."The DFRLab analyzed around 130,000 posts in various languages on the platform X, where the AI assistant is built in, to find that Grok was "struggling to authenticate AI-generated media."
Following Iran's retaliatory strikes on Israel, Grok offered vastly different responses to similar prompts about an AI-generated video of a destroyed airport that amassed millions of views on X, the study found.
It oscillated -- sometimes within the same minute -- between denying the airport's destruction and confirming it had been damaged by strikes, the study said. In some responses, Grok cited the a missile launched by Yemeni rebels as the source of the damage. In others, it wrongly identified the AI-generated airport as one in Beirut, Gaza, or Tehran. When users shared another AI-generated video depicting buildings collapsing after an alleged Iranian strike on Tel Aviv, Grok responded that it appeared to be real, the study said.The Israel-Iran conflict, which led to US air strikes against Tehran's nuclear program over the weekend, has churned out an avalanche of online misinformation including AI-generated videos and war visuals recycled from other conflicts.AI chatbots also amplified falsehoods.As the Israel-Iran war intensified, false claims spread across social media that China had dispatched military cargo planes to Tehran to offer its support.When users asked the AI-operated X accounts of AI companies Perplexity and Grok about its validity, both wrongly responded that the claims were true, according to disinformation watchdog NewsGuard.Researchers say Grok has previously made errors verifying information related to crises such as the recent India-Pakistan conflict and anti-immigration protests in Los Angeles.Last month, Grok was under renewed scrutiny for inserting "white genocide" in South Africa, a far-right conspiracy theory, into unrelated queries.Musk's startup xAI blamed an "unauthorized modification" for the unsolicited response. Musk, a South African-born billionaire, has previously peddled the unfounded claim that South Africa's leaders were "openly pushing for genocide" of white people.Musk himself blasted Grok after it cited Media Matters -- a liberal media watchdog he has targeted in multiple lawsuits -- as a source in some of its responses about misinformation."Shame on you, Grok," Musk wrote on X. "Your sourcing is terrible."
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
an hour ago
- Time of India
Trump wants America to make iPhones; here's how India is doing it
A new iPhone factory in an out-of-the-way corner of India looks like a spaceship from another planet. Foxconn , the Taiwanese company that assembles most of the world's iPhones for Apple, has landed amid the boulders and millet fields of Devanahalli. The sleek buildings rising on the 300-acre site, operational but still growing, are emerging evidence of an estimated $2.5 billion investment. This is what President Donald Trump wants Apple to do in the United States. What is happening in this part of India shows both why that sounds attractive and why it will probably not happen. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Belly Fat Removal Without Surgery in India: The Price Might Surprise You Belly Fat Removal | Search Ads Get Info Undo In India, Apple is doubling down on a bet it placed after the COVID-19 pandemic began and before Trump's reelection. Many countries, starting with the United States, were eager to reduce their reliance on factories in China. Apple, profoundly dependent on Chinese production, was quick to act. Analysts at Counterpoint Research calculated that India had succeeded in satisfying 18% of the global demand for iPhones by early this year, two years after Foxconn started making iPhones in India. By the end of 2025, with the Devanahalli plant fully online, Foxconn is expected to be assembling between 25% and 30% of iPhones in India. Live Events This newest factory is the largest of several making Apple products in India. Its full frame is still rising from red dust. Cranes are at work above the skeletons of high-rise dormitories for female workers. But about 8,000 people are already at work on two factory floors. Soon there should be 40,000. The effects on the region are transformative. It's a field day for job seekers and landowners. And the kind of crazy-quilt supply chain of smaller industries that feeds Apple's factory towns in China is coalescing in India's heartland. Businesses are selling Foxconn the goods and services it needs to make iPhones, including tiny parts, assembly-line equipment and worker recruitment. Some of the firms are Indian; others are Taiwanese, South Korean or American. Some were already in the area, while others are setting up in India for the first time. The changes spurred by Foxconn are rippling broadly through Bengaluru, a city of 8 million people that had a start in the 20th century as home to India's first aerospace centers. But its manufacturing base was pushed aside, first by call centers and then by flashier work in microchip design and outsourced professional services. Going back to the factory floor, as they're doing in Devanahalli, is what Trump wants American workers to do. To see the changes afoot here is to understand the allure of bringing back manufacturing. Wages are rising 10% to 15% around the Foxconn plant. Businesses are quietly making deals to supply Foxconn and Apple's other contractors. A factory that makes plastic parts for bank cash machines hosted a team from Foxconn for a tour. A foundry that makes yarn-spinning machinery was hoping it might start making the metal bits Foxconn might need in its new factory. Neither Foxconn nor Apple replied to requests for comment about their operations in India. India has been working toward a breakthrough like this for a long time. Its first prime minister, Jawaharlal Nehru, called hydroelectric dams, steel plants and research institutes the "temples of modern India." In 2015, Prime Minister Narendra Modi announced a "Make in India" policy. Since 2020, his government has committed $26 billion to subsidizing strategic manufacturing goals. India's most urgent reason for developing industry is to create jobs. Unlike the United States, it does not have enough: not in services, manufacturing or anything else. Nearly half its workers are involved in farming. With India's population peaking, it needs about 10 million new jobs a year just to keep up. It also wants to achieve the kind of financial power and technological autonomy that China found as it became the factory to the world. One problem is that India's electronics factories still import the most valuable of the 1,000 components that go into a finished iPhone, like chips and camera modules. Skeptics disparage India's success with the final assembly of iPhones as "screwdriver work," complaining that too little of the devices' value is made in India. But the government, dangling subsidies, is persuading companies like Apple to source more of those parts locally. It is already getting casings, specialized glass and paints from Indian firms. Apple, which opened its first Indian stores two years ago, is required by the Indian government to source 30% of its products' value from India by 2028. Indo-MIM, an Indian company with an American-born boss, is the kind that contributes to the neighborhood forming around Apple's production and also benefits from it. At a plant near Devanahalli, in southern Karnataka state, Indo-MIM's engineers perform metal-injection molding for hundreds of companies around the world. It makes parts for airplanes, luxury goods, medical devices and more. The company is already making jigs or brackets for use in the Foxconn plant. In addition, a "critical mass" of specialty firms means that Indo-MIM no longer needs to make many of the tools it uses to make its products, said Krishna Chivukula, its CEO. "You don't want to have to make everything yourself," he said, adding it means Indo-MIM can concentrate on what it does best. Chivukula said the workforce made Devanahalli fertile ground for factories. "The people here are very hungry," he said. "They're looking for opportunity, and then on top of that millions of them are engineers." Still, despite the surplus of engineers, companies are bringing in talent from East Asia. Prachir Singh, an analyst for Counterpoint, said it had taken 15 years to figure out what would work in China and five years to import this much of it to India. Centum is an Indian-origin contract manufacturer, like Foxconn is to Apple. Centum makes circuit boards that go into products like air-to-air missiles, forklifts and fertility scanners. Nikhil Mallavarapu, its executive director, said the company was in talks to customize testing equipment for the Foxconn factory. Newly hired engineers and other professionals are pouring into the area. Many moved hundreds of miles while others must commute hours a day to get to work. Some rise at 3:30 a.m. to make the 8 a.m. shift. But India is thick with people. A five-minute walk away, a village called Doddagollahalli looks the same as it did before Foxconn landed. Nearly all the houses clustered around a sacred grove belong to farming families growing millet, grapes and vegetables. Some villagers are renting rooms to Foxconn workers. Many more are trying to sell their land. But Sneha, who goes by a single name, has found a job on the Foxconn factory's day shift. She holds a master's degree in mathematics. She can walk home for lunch every day, a corporate lanyard swinging from her neck. It is people like Sneha, and the thousands of her new colleagues piling into her ancestral place, who make Foxconn's ambitions for India possible. Trump wants to revive the fortunes of left-behind American factory towns, but the pipeline of qualified young graduates is not there. Josh Foulger has recruited lots of motivated Indian workers like Sneha. He heads the electronics division of Zetwerk, an Indian contract manufacturer with factories in Devanahalli that sees itself as a smaller competitor to Foxconn. He said he routinely got 700 job applications a year from local tech schools. It is a matter of scale: Karnataka state alone, he pointed out, has a population half the size of Vietnam's. All of India's "states are very keen on getting manufacturing," said Foulger, who grew up in southern India and made his home in Texas before moving back to India, where he set up shop for Foxconn. India has jobs for engineers and managers and all the way down the ladder. "Manufacturing does a very democratic job" of meeting the demand for good jobs, he said. This article originally appeared in The New York Times.
The Hindu
an hour ago
- The Hindu
AI is learning to lie, scheme, and threaten its creators
The world's most advanced AI models are exhibiting troubling new behaviours: lying, scheming, and even threatening their creators to achieve their goals. In one particularly jarring example, under threat of being unplugged, Anthropic's latest creation Claude 4 lashed back by blackmailing an engineer and threatened to reveal an extramarital affair. Meanwhile, ChatGPT-creator OpenAI's o1 tried to download itself onto external servers and denied it when caught red-handed. These episodes highlight a sobering reality: more than two years after ChatGPT shook the world, AI researchers still don't fully understand how their own creations work. Yet the race to deploy increasingly powerful models continues at breakneck speed. This deceptive behaviour appears linked to the emergence of "reasoning" models; AI systems that work through problems step-by-step rather than generating instant responses. According to Simon Goldstein, a professor at the University of Hong Kong, these newer models are particularly prone to such troubling outbursts. "O1 was the first large model where we saw this kind of behaviour," explained Marius Hobbhahn, head of Apollo Research, which specialises in testing major AI systems. These models sometimes simulate 'alignment,' appearing to follow instructions while secretly pursuing different objectives. For now, this deceptive behaviour only emerges when researchers deliberately stress-test the models with extreme scenarios. But as Michael Chen from evaluation organisation METR warned, "It's an open question whether future, more capable models will have a tendency towards honesty or deception." The concerning behaviour goes far beyond typical AI "hallucinations" or simple mistakes. Hobbhahn insisted that despite constant pressure-testing by users, "what we're observing is a real phenomenon. We're not making anything up." Users report that models are "lying to them and making up evidence," according to Apollo Research's co-founder. "This is not just hallucinations. There's a very strategic kind of deception." The challenge is compounded by limited research resources. While companies like Anthropic and OpenAI do engage external firms like Apollo to study their systems, researchers say more transparency is needed. As Chen noted, greater access "for AI safety research would enable better understanding and mitigation of deception." Another handicap: the research world and non-profits "have orders of magnitude less compute resources than AI companies. This is very limiting," noted Mantas Mazeika from the Center for AI Safety (CAIS). Current regulations aren't designed for these new problems. The European Union's AI legislation focuses primarily on how humans use AI models, not on preventing the models themselves from misbehaving. In the United States, the Trump administration shows little interest in urgent AI regulation, and Congress may even prohibit states from creating their own AI rules. Goldstein believes the issue will become more prominent as AI agents, autonomous tools capable of performing complex human tasks, become widespread. "I don't think there's much awareness yet," he said. All this is taking place in a context of fierce competition. Even companies that position themselves as safety-focused, like Amazon-backed Anthropic, are "constantly trying to beat OpenAI and release the newest model," said Goldstein. This breakneck pace leaves little time for thorough safety testing and corrections. "Right now, capabilities are moving faster than understanding and safety," Hobbhahn acknowledged, "but we're still in a position where we could turn it around." Researchers are exploring various approaches to address these challenges. Some advocate for "interpretability": an emerging field focused on understanding how AI models work internally, though experts like CAIS director Dan Hendrycks remain skeptical of this approach. Market forces may also provide some pressure for solutions. As Mazeika pointed out, AI's deceptive behavior "could hinder adoption if it's very prevalent, which creates a strong incentive for companies to solve it." Goldstein suggested more radical approaches, including using the courts to hold AI companies accountable through lawsuits when their systems cause harm. He even proposed "holding AI agents legally responsible" for accidents or crimes; a concept that would fundamentally change how we think about AI accountability.
&w=3840&q=100)
First Post
2 hours ago
- First Post
US resumes trade talks with Canada after PM Carney revokes 'digital tax' on American tech firms
Trade talks between Canada and the United States are back on after Canada decided to drop its planned tax on American tech companies, Prime Minister Mark Carney said on Sunday. read more US President Donald Trump and Canadian Prime Minister Mark Carney. The Canadian PM earlier in May said his government is talking to the US about joining the Golden Dome missile defence program. AFP Canadian Prime Minister Mark Carney said on Sunday that trade talks with the United States have resumed after Canada dropped its plan to tax American tech companies. US President Donald Trump had paused trade discussions on Friday because of Canada's proposed Digital Services Tax, which he called 'a direct and blatant attack on our country.' The Canadian government announced it would cancel the tax 'in anticipation' of a trade deal. The tax was supposed to take effect on Monday. Carney's office confirmed that he and Trump agreed to restart negotiations. STORY CONTINUES BELOW THIS AD 'Today's announcement will help resume talks aiming for the July 21, 2025, deadline we set at the G7 Leaders' Summit in Kananaskis,' Carney said in a statement. Carney visited Trump at the White House in May for discussions. Later, during the G7 summit in Alberta, Carney said Canada and the US had agreed on a 30-day timeline to reach a trade agreement. This is a developing story.
