WASHINGTON DC: The rapid rise of artificial intelligence is reshaping the tech industry, driving growth but raising security and investment concerns. Open-source large language models (LLMs) have become a target for hackers, researchers warn, as unsecured models can be exploited for spam, phishing, disinformation, and other malicious activities.

A joint study by SentinelOne and Censys found that thousands of LLM deployments, including variants of Meta’s Llama and Google DeepMind’s Gemma, are exposed online, with hundreds having removed safety guardrails.

About 30% of these hosts operate from China and 20% from the US. Experts stress that responsibility for misuse is shared across developers, deployers, and the wider AI ecosystem.

Meanwhile, Big Tech earnings reveal investor focus on AI returns. Meta saw a 24% revenue rise in Q4 2025, boosted by AI-driven ad targeting, while Microsoft’s Azure cloud growth disappointed despite heavy investment in OpenAI.

Tesla reported its first annual revenue decline as it pivots to AI, robotics, and humanoid robots, planning a USD 20 billion capital outlay.

Analysts note a growing gap between AI ambitions and investor patience for returns, highlighting the balance between innovation, profitability, and security in the AI era.