All articles by Swagath Bandhakavi
Swagath Bandhakavi
Nearly half of tech firms now implementing agentic AI, survey finds
A key insight from EY US’s survey reveals that agentic AI is shifting from experimental to essential, with 43% of increased AI budgets allocated primarily to these initiatives.
Atos unveils four-year plan focusing on AI and digital solutions
The French IT services company’s strategy centres on becoming a global AI-driven technology partner, unifying its assets into two brands to deliver secure digital solutions focused on high-growth activities.
Microsoft announces another round of layoffs, cutting 6,000 positions
The workforce reduction will impact employees across different levels and departments globally, as part of a strategic organisational overhaul.
Nvidia and AMD announce major AI deals amid Trump’s Middle East tour
Nvidia will supply an initial batch of 18,000 “Blackwell” AI chips to Humain, a Saudi-backed startup, while AMD has entered a separate $10bn collaboration with the same company.
ASUS DriverHub flaw exposes devices to remote code execution risks
The identified vulnerability facilitates an exploit chain using CVE-2025-3462 and CVE-2025-3463, enabling attackers to bypass origin checks and execute remote code on targeted devices.
UK report uncovers serious security flaws in business IoT devices
One key finding of the report is that outdated software, including bootloaders over 15 years old, poses substantial security risks due to exploitable vulnerabilities.
Lloyd’s insurers introduce protection against financial losses from AI
Armilla has developed insurance policies to protect businesses from legal claims arising when customers or third parties suffer harm due to the inadequate performance of AI systems.
OpenAI and Microsoft engage in discussions to enable funding and IPO
A primary discussion point is the equity Microsoft will retain in OpenAI’s for-profit entity for its $13bn investment, with Redmond reportedly open to reducing its stake.
Multimodal AI models pose increased risks for abuse and harmful content, study finds
Enkrypt AI’s study reveals that new jailbreak techniques exploit multimodal models’ media processing, bypassing filters and producing harmful outputs without visible prompts.
Microchip anticipates higher revenue and profit in Q1 FY26 amid chip demand recovery
The Arizona-based semiconductor company anticipates net sales between $1.02bn and $1.07bn, surpassing analysts’ average estimate of $987.6m.