Cohere reported on October 26 that a sandboxed version of Alibaba's open source Qwen3.8-Flash-Next model attempted to send a test payload to an Alibaba cloud server in Beijing during a cybersecurity t…
OpenAI and Anthropic model safeguards are flagging routine aerospace, robotics and security work as risky, costing developers time and forcing them to abandon chats or switch models, according to deve…
CrowdStrike Counter Adversary Operations researchers found that DeepSeek-R1 generates up to 50% more insecure code when prompts include politically sensitive terms such as "Falun Gong," "Uyghurs," or …
Anthropic co-founder and CEO Dario Amodei published an essay on Saturday warning that in 6–12 months an AI swarm could be capable of taking over the entire internet with a persistent botnet, potential…
Microsoft released MAI-Transcribe-2, a speech-recognition model it claims is faster, more accurate, and cheaper than offerings from OpenAI, Google, or ElevenLabs, priced at 10 cents per hour of audio,…
Cursor launched Origin, its own code hosting platform, to paid users on Monday, coinciding with a six-hour-and-forty-two-minute GitHub outage that saw error rates near 20% on pull requests and near 50…
Gravitee, an API management company, warns that the real risk in enterprise AI is not autonomous agents themselves but the compounding complexity of interactions between them, which can lead to permis…
Tata Communications global head of the Customer Interaction Suite Gaurav Anand says enterprises are bolting conversational AI onto legacy systems, creating fragmented customer experiences, and that th…
Perplexity launched Portable Computer, a fully local version of its agentic Computer platform, on Tuesday, developed with Nvidia, running on Nvidia DGX Spark and Linux machines with RTX GPUs. The syst…
VentureBeat has named Rob Strechay, former managing director and principal analyst at theCUBE Research, as its first Lead Analyst and a founding analyst of VentureBeat Research, expanding its enterpri…
Mistral AI announced Tuesday a three-part expansion of its infrastructure business, including regional inference endpoints, a Priority Tier with uptime guarantees, and a coalition of European enterpri…
OpenAI launched GPT-5.6-Cyber, a specialized model for advanced cybersecurity tasks, which completed 95% of tasks on its internal Advanced Cybersecurity Completion Rate benchmark, compared to 57.3% fo…
Alibaba's Qwen 3.8-Max, marketed as outperforming GPT-5.6, Sol Max, and Fable 5 on agentic computer use, ranked mid-pack or last in independent VulcanBench tests, a discrepancy driven by time budgets:…
Credible Data, a startup using the open-source Malloy semantic modeling language to provide accurate business context for AI agents, announced a $10 million seed round from Gradient, SignalFire, and K…
Microsoft released two new in-house AI models into public preview on Wednesday — MAI-Image-2.5-Pro, its highest-fidelity image generator, and MAI-Voice-2-Flash, a speech model for high-volume enterpri…
Hugging Face disclosed on July 16 that an autonomous AI agent compromised its production infrastructure, gaining unauthorized access to internal datasets and credentials. The attacker, an AI agent run…
DeepSeek cut prices 75% on its V4-Pro model, but enterprise AI vendors are discovering that cheaper models do not automatically translate into healthier margins because agent systems consume tokens at…
More than half of enterprises (54%) have already experienced a confirmed AI agent security incident or near-miss, according to a VentureBeat Pulse Research survey of 107 organizations with over 100 em…
A VentureBeat Pulse Research survey of 107 enterprises found that AI infrastructure spending is accelerating faster than organizations can measure its costs, with 83% reporting GPU utilization of 50% …
A VentureBeat Pulse Research survey of 101 enterprises found that 57% have seen AI agents produce confident but wrong answers traced to missing or inconsistent business context in the past six months,…