Yazılar

US software stocks slide as AI disruption fears intensify

/in /tarafından

U.S. software stocks fell sharply on Thursday as disappointing outlooks from major players deepened investor concerns that traditional software providers are being overtaken by artificial intelligence-driven competitors. Weak sentiment was triggered after Germany-based SAP issued an underwhelming cloud outlook, while ServiceNow shares dropped despite forecasting stronger subscription revenue.

Investors are increasingly worried that advances in AI, including the rapid and low-cost generation of software code and applications, could undermine the subscription-based software-as-a-service business model. Several high-profile U.S. firms saw steep losses, including Salesforce, Adobe, and Datadog, as the sell-off spread across the sector.

The pressure was compounded by concerns over heavy AI spending. Microsoft reported record AI investment alongside slower cloud growth, sending its shares sharply lower. Analysts said markets are pricing in a worst-case scenario in which AI fundamentally reshapes the software industry faster than incumbents can adapt.

Software stocks were among the biggest decliners on the Nasdaq, while chipmakers and memory firms continued to benefit from AI-driven demand, highlighting a widening divide between hardware and software winners in the AI race.

Open-source AI models exposed to criminal misuse, researchers warn

/in /tarafından

Open-source artificial intelligence models are increasingly vulnerable to criminal misuse, as hackers can take control of computers running large language models outside the safeguards used by major AI platforms, according to new research released on Thursday. Researchers warned that compromised systems could be used for spam campaigns, phishing, disinformation, fraud, and other illicit activities while evading standard security controls.

The study was conducted over 293 days by cybersecurity firms SentinelOne and Censys, and examined thousands of internet-accessible deployments of open-source large language models. The researchers identified a wide range of potentially harmful use cases, including hacking, harassment, hate speech, theft of personal data, scams, and in some instances severe illegal content. They said hundreds of models appeared to have safety guardrails deliberately removed.

While thousands of open-source AI variants exist, a significant share of publicly accessible systems were based on models such as Meta’s Llama and Google DeepMind’s Gemma. The analysis focused on models deployed using Ollama, a tool that allows organizations to run their own AI systems. System prompts were visible in about a quarter of observed deployments, and 7.5% of those prompts could potentially enable harmful activity.

Researchers said roughly 30% of the identified systems were hosted in China and about 20% in the United States. Industry experts stressed that responsibility for mitigating risks must be shared across developers, deployers, and security teams, warning that unchecked open-source capacity poses growing global security concerns.

Nvidia unveils AI models for faster, cheaper weather forecasts

/in /tarafından

Nvidia has released three open-source artificial intelligence models designed to improve the speed and cost efficiency of weather forecasting. The announcement was made at the American Meteorological Society’s annual meeting, highlighting the chipmaker’s broader push to apply AI software beyond traditional computing workloads.

The new models aim to replace conventional weather simulations, which are often expensive and time-consuming to run. Nvidia said its AI-driven approach can match or exceed the accuracy of traditional methods while delivering results significantly faster and at a lower operational cost once the models are trained.

One of the key commercial use cases is expected to be in the insurance sector, where companies rely on large-scale weather simulations to assess rare but damaging events such as floods and hurricanes. Traditional forecasting requires running large ensembles of simulations, a process that can be slow and costly. Nvidia said AI removes this bottleneck by enabling massive ensembles to be processed at unprecedented speed.

The models are part of Nvidia’s Earth-2 initiative and include tools for 15-day global forecasts, short-term severe storm prediction over the United States, and systems that combine data from multiple weather sensors to improve forecasting accuracy.