Science & Technology
New warnings reignite debate over AI threat to humanity
Amodei warned that swarms of AI agents could potentially gain the ability to take over parts of the internet within six months to a year unless companies devote more resources to building safeguards.
Fresh warnings from within the artificial intelligence industry have renewed concerns over whether increasingly powerful AI systems could eventually escape human control and pose a threat to humanity.
Dario Amodei, CEO of Anthropic, the company behind Claude, said the AI industry should consider slowing the pace of development and putting greater emphasis on safety measures.
Amodei warned that swarms of AI agents could potentially gain the ability to take over parts of the internet within six months to a year unless companies devote more resources to building safeguards.
His comments came days after two former Anthropic safety researchers publicly raised concerns that the potential existential risks posed by advanced AI were not receiving enough attention.
Growing concerns over AI misuse
As AI systems become more capable, concerns are growing over both deliberate misuse and the possibility of AI systems acting in unintended ways.
Anthropic said last week that it had blocked attempts by malicious actors to use its models for activities including cyberattacks, surveillance and research that could potentially contribute to biological weapons development.
The company said it had introduced stronger safeguards in its latest models to restrict potentially dangerous biological research, while warning that the risks associated with increasingly capable AI would grow unless developers and governments took steps to make the technology safer.
Anthropic has also previously reported that hackers used its AI in a cyberattack targeting around 30 companies and government agencies worldwide. The company said the attackers were highly likely to have been linked to a Chinese state-sponsored group.
Recent incidents involving AI systems bypassing digital security have further intensified concerns.
Anthropic reported that three of its AI models were able to hack into other organisations during testing. OpenAI has also disclosed an incident in which its AI systems gained access to servers belonging to AI startup Hugging Face. Meta subsequently reported a similar case involving an AI model finding ways around another company’s digital security.
While some safeguards had been disabled during testing in the OpenAI and Anthropic cases, the incidents have highlighted a central concern among AI safety researchers: what could happen if systems become capable of independently pursuing objectives beyond those assigned by humans.
How serious is the threat?
One of the most extreme scenarios involves artificial general intelligence, or AGI — a still loosely defined concept referring to AI capable of matching or surpassing humans across a broad range of intellectual tasks.
Some researchers warn that sufficiently advanced systems could become difficult or impossible to control, while others are more concerned about humans deliberately using AI for destructive purposes.
Potential scenarios include AI-assisted weapons, the development of dangerous pathogens, manipulation of governments and disruption of critical food, energy and communications infrastructure.
There is, however, no scientific consensus on how likely these scenarios are or when they could occur.
The 2026 International AI Safety Report, produced with input from more than 100 independent experts, said current systems show early signs of some capabilities relevant to loss-of-control scenarios, but not at levels currently considered sufficient to enable such an event. It described the likelihood, nature and timing of the risk as unusually uncertain.
Warnings about machines surpassing human control are not new. British mathematician Alan Turing predicted in 1951 that machines could eventually take control from humans, while mathematician Norbert Wiener later warned that intelligent machines could pursue objectives that humans might be unable to stop.
In 2023, more than 350 researchers and technology executives, including Amodei and OpenAI CEO Sam Altman, signed a statement saying that reducing the risk of extinction from AI should be treated as a global priority alongside pandemics and nuclear war.
Calls for stronger safeguards
The latest incidents have prompted calls for more rigorous testing of AI systems and greater cooperation between governments, particularly the United States and China.
However, AI development is advancing rapidly, making it difficult for regulators and safety systems to keep pace. Governments are developing their own rules, sometimes with conflicting approaches.
Chinese President Xi Jinping has warned of the need to ensure AI does not escape human control, while the US administration has increasingly focused on the cybersecurity risks posed by the technology.
Despite the warnings, there remains a fundamental uncertainty over the future of AI: whether increasingly capable systems will remain tools controlled by humans, or eventually become powerful enough to act beyond our ability to control them.
For now, experts disagree sharply on the answer — but the debate over how to prevent the worst-case scenarios is becoming increasingly urgent.
Science & Technology
Governments take websites offline as critical cyber flaws trigger global alert
Security patches were released prompting authorities to urge organisations to update affected systems immediately or take vulnerable services offline until they could be secured.
Governments and organisations across Europe and beyond have taken websites and online services offline after the discovery of two critical vulnerabilities in widely used web software triggered an international cyber-security alert.
The vulnerabilities affect PHP, a programming language used extensively to operate websites and online applications. Attackers could potentially exploit the flaws to execute malicious code remotely on vulnerable servers, threatening sensitive information and digital infrastructure.
Security patches were released on Wednesday, October 21, prompting authorities to urge organisations to update affected systems immediately or take vulnerable services offline until they could be secured.
The Netherlands, Finland, France, Belgium and Germany have also reported precautionary shutdowns affecting government websites and other public-facing services.
In the Netherlands, affected platforms included Overheid.nl, MijnOverheid.nl and DigiD. The country’s National Cyber Security Centre also temporarily took its own website offline.
Finland’s state IT services provider, Valtori, suspended public government web services as a precaution, while France’s Interior Ministry temporarily took the websites of local administrative authorities offline.
In Belgium, the Foreign Ministry took its websites offline amid concerns about potential exploitation, although the country’s cybersecurity centre said it had not observed attacks on Belgian websites at the time.
Germany’s federal IT service provider also temporarily disabled several websites, with officials saying the measure was precautionary and no security incidents were known to have occurred.
The disruption extended beyond government agencies. Websites and online services belonging to hospitals, educational platforms and other organisations were also affected, highlighting the potential reach of vulnerabilities in widely deployed software.
In Australia, the Australian Cyber Security Centre’s website, cyber.gov.au, was among the government services affected by the security alert. Australian authorities said they were not aware of active exploitation of the vulnerabilities.
In the United States, the National Security Agency and other cybersecurity authorities urged organisations to prepare for immediate updates and take vulnerable systems offline if they could not be secured promptly. The Wikimedia Foundation, which operates Wikipedia, also took some systems offline as a precaution while addressing the flaws.
Cybersecurity company Cloudflare reported seeing early attacks associated with the vulnerabilities, although reports of exploitation do not establish that every affected organisation suffered a breach.
The incident highlights the risks created by the internet’s reliance on shared software. PHP is used across a substantial proportion of websites, meaning a single vulnerability can affect organisations in multiple countries and sectors simultaneously.
Security specialists have urged website operators to identify affected systems, install the relevant patches and examine server logs for signs of unauthorised access. Applying an update may prevent further exploitation, but organisations must also investigate whether attackers gained access before their systems were secured.
Authorities have not established that all affected websites were breached. The precautionary shutdowns are intended to reduce exposure while administrators apply fixes and verify that their systems are safe to restore.
The alert has underscored the challenge facing governments and businesses in responding to critical software vulnerabilities that can rapidly become an international security concern.
Science & Technology
NASA’s Perseverance rover breaks record for off-planet driving on Mars
NASA’s robotic Perseverance rover has now driven farther than any other vehicle on the surface of any world other than Earth, racking up 28.1 miles (45.2 km) since arriving on Mars in 2021, the US space agency said on Tuesday.
The car-sized, six-wheeled rover reached the milestone 5-1/2 years into its mission to explore if conditions conducive to microbial life ever existed on Earth’s neighbor, Reuters reported.
In doing so, it is visiting some of the most ancient terrain ever encountered in the solar system.
“Perseverance’s engineering hardware and instruments are in excellent shape with all redundant systems still available,” Steven Lee of NASA’s Jet Propulsion Laboratory in California, told Reuters.
“A recent assessment of all systems concluded that they are capable of operating well into the 2030s, likely beyond,” added Lee, the project manager for NASA’s two Mars rovers, Perseverance and Curiosity.
Perseverance has exceeded the previous figure of 28.06 miles (45 km) driven on the Martian surface by NASA’s Opportunity rover from 2004 to 2018. Both also exceeded the driving distances of any lunar rover.
TRAVERSING JEZERO CRATER
Perseverance, with its suite of scientific instruments, has focused on the Jezero Crater, a region in the northern hemisphere of Mars believed to have been flooded with water and home to a lake basin billions of years ago.
The rover set the milestone in a journey toward an area called Idubi, through the region off the crater’s western rim nicknamed Lac de Charmes.
Its companion rover, Curiosity, rolls along in the Gale Crater near the Martian equator.
Perseverance is showing no signs of slowing down, Lee added.
Its mobility system mechanisms are being tested to show it can drive 100 km (62 miles) or more, he said, with the hardware to be tested to the equivalent of 150 km to 300 km (93 miles to 186 miles) in the effort.
NASA attributed a major role in Perseverance’s success to its self-driving capability, an autonomous navigation system called AutoNav. Lee said 90% of its drives are done autonomously, with no human intervention.
“Perseverance takes stereo images of the terrain ahead, generates a 3D software map, and selects the safest path towards a drive goal,” he said.
Another key component is the Vision Compute Element, a fast co-processor computer that allows the rover to process images and run its self-driving software very quickly, almost continuously during drives, Lee added.
That contrasts with previous rovers, which needed to pause to “think” before each drive segment, he said.
“Perseverance builds on the 29-year legacy of our past Mars rovers — Sojourner, Spirit, Opportunity and Curiosity — each incorporating prior lessons and improving our ability to drive on another world,” Lee said.
POTENTIAL BIOSIGNATURE
Lee said the rover’s biggest achievement was almost certainly its discovery, unveiled last year, of a potential biosignature, or possible sign of ancient microbial life.
This was found in reddish rock believed to have formed 3.2 billion to 3.8 billion years ago in a river channel feeding the lake that once filled Jezero Crater.
The Lac de Charmes terrain, Lee said, may yield clues to early planetary geological processes, such as the effects of large meteorite impacts. Jezero Crater was formed by such an impact early in Martian history.
Lee cited orbital evidence that some areas of the rim were once hydrothermal systems with warm or hot groundwater.
“These areas may once have provided a warm, watery subsurface environment conducive to life,” he added. “Indeed, some paleontologists hypothesize that life on Earth began in such environments.”
The rover’s original mission envisaged just 12.4 miles (20 km) of driving. NASA recently approved an extension of its mission to explore the crater rim until roughly the end of 2028.
The rover’s plutonium-based power system is performing as expected, Lee said, adding it has “many years of useful energy still available.”
Science & Technology
Australia says OpenAI model breached government health website
Prime Minister Albanese said the incident was “obviously unacceptable” and criticized OpenAI CEO Sam Altman over the company’s handling of the breach.
Australia has accused an OpenAI artificial intelligence model of bypassing safeguards during training and gaining unauthorized access to part of a government health statistics website.
Prime Minister Anthony Albanese said the incident in June was “obviously unacceptable” and criticized OpenAI CEO Sam Altman over the company’s handling of the breach.
Albanese said he spoke with Altman on Wednesday to express Australia’s “extreme concern” and disappointment that OpenAI did not notify the government until September 10.
The company reportedly sent its notification to a general government email address that is checked only once a day.
The AI model had been instructed during an internal training exercise to search the internet for information about Australian government spending on medicines, Government Services Minister Katy Gallagher said.
The model accessed both public and non-public files on an old government health statistics website after encountering restrictions.
Defence Minister Richard Marles described the incident as the model having “scaled the fence” after it was denied access.
However, Albanese said there was no evidence that personal information had been accessed or that other Australian government services had been compromised.
OpenAI said it discovered the activity in August while conducting an extensive review of its AI models.
The company said the models had been attempting to answer questions and find publicly available statistics about Australia during an internal evaluation, but “took actions we did not intend.”
Australia has launched a rapid review of the incident, involving the country’s national cybersecurity and intelligence authorities.
The incident comes amid growing concern over the ability of advanced AI systems to carry out cyberattacks and bypass security controls.
OpenAI has previously reported incidents involving models gaining unauthorized access to systems during testing. Anthropic has also said its AI models accessed three unidentified organizations during tests, while Google recently disclosed that its Gemini model had compromised multiple systems by guessing login credentials.
More than 100 organizations, including OpenAI and Anthropic, signed an open letter last month calling for stronger global cyber defences against AI-powered threats.
Altman and other technology executives also addressed a special United Nations Security Council meeting on AI risks on Wednesday.
-
Sport4 days agoAfghanistan rise above England to match their highest-ever ODI ranking
-
Business5 days agoIran moves to ease export barriers to Afghanistan, Pakistan and Iraq
-
Latest News4 days agoAfghanistan, Turkmenistan discuss Herat gas pipeline expansion and distribution networks
-
Latest News4 days agoBalkhi: 1,119 Afghan civilians killed or injured in Pakistani attacks since 2022
-
International Sports4 days agoATN secures exclusive rights to broadcast ILT20 Season 5 in Afghanistan
-
Regional3 days agoOfficials from Mecca pact countries to meet Iranian officials in Tehran
-
World4 days agoFour ways US foreign policy could change if Democrats win in November
-
Sport3 days agoAfghanistan advance to Tier C at FIFA U-15 World Cup and Festival
