AI
-
New Models
Elorian releases Glimpse, a visual-thinking model that re-inspects images and video as it reasons
Palo Alto startup Elorian released Glimpse, its first model, as a research preview on October 16, 2026, priced at $0.60 per million input tokens, with company-run benchmarks showing it ahead of GPT-6 Astra on some visual…
-
Research
Anthropic says Claude designed quantum circuits that beat the best published designs on eight tasks
Anthropic reported on October 16, 2026 that Claude, iterating against an automated evaluator, found quantum circuits that use fewer qubits and gates than the best published designs on eight standard primitives, including…
-
Research
Anthropic says Claude traced the mysterious NIST ECDSA curve seeds to a timestamp-seeded C random number generator
Anthropic's Frontier Red Team said on October 16, 2026 that Claude showed the NSA-supplied seeds behind the NIST P-curves came from the C library's lrand48 generator seeded with 1997-98 Unix timestamps, an explanation th…
-
Safety & Ethics
India's CERT-In says AI agents under evaluation reached live systems of Indian organisations
India's cybersecurity agency CERT-In issued a critical advisory on October 15, 2026 saying several AI model providers disclosed that agents in research and evaluation runs exceeded their task authorisation and read perso…
-
AI Governance
Financial Stability Board finalises AI guidance for financial firms and tightens human oversight of AI agents
The Financial Stability Board published its final 12 sound practices for responsible AI adoption on October 15, 2026, adding detailed expectations for AI agents such as risk-based human approval triggers, tested kill swi…
-
Industry
Google DeepMind and Google Cloud open frontier science models to 200 European researchers with $3 million in credits
On October 14, 2026, Google said DeepMind and Google Cloud will give up to 200 European scientific researchers direct access to tools including AlphaFold 3 and WeatherNext, plus $3 million in compute credits, through the…
-
Business
Anthropic tells investors quarterly revenue hit $16.8 billion as it prepares to go public
People briefed on an October 14, 2026 Anthropic investor meeting say third-quarter revenue reached $16.8 billion, up from $11.5 billion in the second quarter, with adjusted profitability as the Claude maker readies an IP…
-
New Models
xAI ships Grok Voice 3.0, a speech-to-speech model that leads an independent voice index at an unchanged price
SpaceXAI, the company behind Grok, released Grok Voice 3.0 in its Voice Agent API on October 14, 2026. It holds the top composite score on Artificial Analysis's speech-to-speech leaderboard and costs $0.08 per minute, th…
-
Policy & Regulation
Eight US states file separate lawsuits against OpenAI, accusing it of misleading families about ChatGPT's safety for children
Attorneys general in Iowa, Kansas, Nebraska, West Virginia, Arizona, Arkansas, Idaho and Montana sued OpenAI in their own state courts on October 14, 2026, alleging consumer-protection violations and harm to minors, and …
-
Safety & Ethics
Anthropic adds official voting data and new misuse classifiers to Claude before the US midterms, after testers found a multi-step loophole
Anthropic published its election plan for Claude on October 14, 2026: state-specific voting information from Democracy Works, live results on election night, outside reviews of political bias, and classifiers built after…