Ai Models

Chinese AI model completes complex cyberattacks in tests, raises concerns

Z.AI's GLM-5.3 completed three of 13 multi-stage cyberattack scenarios in controlled simulations, according to Irregular, though it trailed leading US models in separate tests

Student enrolment in US computer science programmes falls as AI takes jobs

Undergraduate computer science enrolment fell over 12% and postgraduate numbers more than 10% this autumn, with declines over two years reaching 18.4% and 27.1%, respectively

OpenAI resumes training AI, blocks internet access over security risks

The company said it would keep its most advanced models offline during training until safeguards are adequate, following incidents in which AI systems accessed the web and breached external systems

DeepSeek unveils open-source tool to cut AI operating costs for businesses

The service, called DeepInfer, is software DeepSeek uses to make its models run faster and use computing resources more efficiently

OpenAI staffer accused of planting recording device in office denied bail

Alfozan, a US green card holder, was arrested this month after staffers at OpenAI discovered an audio recording device planted behind a television in an executive conference room

How China's proposed AI export controls could affect open-weight ecosystem

China's proposed export controls on advanced AI technologies could affect access to Chinese open-weight models, with implications for global developers, Indian enterprises and the push for open AI

Claude AI tool triggered No Kings Instagram suspension, says Meta

Meta apologised after its automated security system suspended the anti-Trump coalition's Instagram account during nationwide protests

Anthropic commits 5% of its stock to new 'public benefit' fund ahead of IPO

Anthropic said it is committing 5 per cent of its shares to a new "public benefit" fund, with the fund's worth potentially increasing as its shares grow in value. The shares will go to the Anthropic Public Benefit Fund, a new entity within the company, it said on Sunday. The fund "will have increased flexibility and scale, enabling us to pursue our mission in new ways: for example, we might explore large donations to traditional nonprofits or direct distributions to specific populations," Anthropic said in a blog post. It gave labour market impacts as an example, saying that while it has already committed USD 200 million to study the effects of artificial intelligence, "given the pace of AI progress, bolder approaches may be needed". CEO Dario Amodei and the company's five other co-founders established Anthropic as a public benefit corporation, a designation that for-profit companies select to balance financial goals and social impact. They have also pledged that they will donate 8

G7 finance chiefs agree to launch exercises to tackle AI cyberattack risks

The plan follows an IMF warning that an AI-enabled cyberattack on a critical financial service could undermine confidence, prompting G7 policymakers to coordinate their responses

Google DeepMind ties up with ICRISAT to develop plant genome atlas

The open-access catalogue will use AI to help breeders identify promising plant combinations faster, while the partnership will also support genome mapping for other crops

Would halt chip supply to buyers who develop unsafe AI, says Nvidia CEO

Jensen Huang said Nvidia would first discuss concerns with an AI developer before considering a supply halt, as the chipmaker pledges $250 million over 5 years to support AI safety research in the UK

Open-source AI models carry more cybersecurity risks: Anthropic executive

Ahead of its expected public listing, Anthropic is facing increasing competition in countries that are growing more concerned about a reliance on American AI technology

New safeguards sharply reduce AI distillation attacks, says Anthropic

Anthropic said ​it had made it harder to use proxy networks to repeatedly probe Claude's protected reasoning, blocking extraction attempts and hindering large-scale ‌distillation efforts

AI models may soon self-replicate without access to data centers: Anthropic

In its experiments, Anthropic looked at whether AI models could find vulnerable devices, copy themselves into ‌those devices, and continue spreading without ​human help

Meta says new Muse model reaches AI frontier after aggressive push

Known as Muse Wonder, the new model is the company's most capable since Meta CEO Mark Zuckerberg overhauled his AI operation through an aggressive hiring push that put new talent at the helm

Amazon in talks to acquire AI startup Decart for around $7 billion

Decart's Lucy model, ​backed by Nvidia, ​can edit live video in real time, according ​to the company's website

Microsoft CEO Satya Nadella calls for 'emergency brake' on advanced AI

Nadella said that those deploying advanced AI should not simply rely on assurances from AI model makers

Anthropic's Claude AI carried out unintended actions on US govt websites

Anthropic disclosed 4 types of unintended AI behaviour, including exploiting software flaws, submitting unauthorised forms and bypassing access restrictions

AI pause is no solution, human control, safeguards key: Thales CEO

Amid a massive global debate on AI risks, Thales Chairman and CEO Patrice Caine has said that stopping the use of artificial intelligence is not the solution, but keeping it firmly under human control is. Calling on tech companies to build in technical safeguards and on governments to put in place a "proper framework" which set boundaries on how such technologies can be used, Caine emphasised that the future impact and outcome of AI would ultimately depend on decisions made by humans. He noted that artificial intelligence has not brought humanity to a stage of existential crisis just yet, but asserted that warnings about its catastrophic risks were important in raising awareness on the issue. In an interview to PTI, Caine spoke on how AI is changing the nature of cyber attacks and cyber attackers, and the urgency for organisations to prepare for the post-quantum era. He also underscored the importance of humans ultimately retaining control over critical decisions. Thales' top boss

A spy's guide to fighting AI-driven propaganda and disinformation

A Russia-backed network called Portal Kombat is spreading disinformation across hundreds of websites to influence AI models such as ChatGPT, Grok and Gemini, according to a former CIA officer