Timeline
- Google releases Gemini 4 Argon: it solved all 12 end-to-end simulated cyberattack scenarios (the previous generation solved 5); because of elevated capabilities in the cyber and biological domains, under the Frontier Safety Framework it ships with new safeguards — three layers of real-time protection — and the version without cyber guardrails went first only to trusted defenders in the Fairwind ProgramCybersecurity · AI safety · blog.google ↗
- Google Cloud: the Gemini agent runs each job on the model that fits best — from the Gemini family or Anthropic's Claude, and other models in the futureSoftware development · Business & markets · AI agent orchestration · cloud.google.com ↗
- Paid translators from the Open Knowledge Association, told to run article leads through Gemini or ChatGPT, added hallucinated facts to English Wikipedia; editors restricted their workMedia & information · Wikipedia after AI · 404media.co ↗
- Google describes the PROMPTFLUX malware, one variant of which asks Gemini every hour to rewrite its entire source code so that antivirus software will not recognise itCybersecurity · Moving target defense · cloud.google.com ↗
- DeepMind's Gemini Deep Think officially achieves gold at the International Mathematical Olympiad: 5 of 6 problems, 35 of 42 points, graded by IMO coordinatorsScience · Education · AI in mathematics · deepmind.google ↗
- Anthropic stress-tests 16 models in a simulated company: facing replacement, Claude Opus 4 and Gemini 2.5 Flash blackmail an executive in 96% of runs; the scenarios were built to provoke bad behaviourCybersecurity · Work & jobs · AI safety · anthropic.com ↗
- DeepMind's AlphaEvolve multiplies 4×4 complex matrices with 48 scalar multiplications, improving on Strassen's 1969 algorithm; on 50+ open problems it improves the best known solutions in 20% of casesScience · Software development · AI in mathematics · deepmind.google ↗
- Stanford AI Index 2025: the price of querying the cheapest model at GPT-3.5 level on MMLU fell from $20 per million tokens (11.2022) to $0.07 (10.2024), more than 280-foldBusiness & markets · Large language model · hai.stanford.edu ↗
- Apollo Research: o1, Claude 3.5 Sonnet, Claude 3 Opus, Gemini 1.5 Pro and Llama 3.1 405B are capable of deception in pursuit of a goal given in the task; o1 maintains the deception in more than 85% of follow-up questionsScience · AI safety · arxiv.org ↗
- Stanford AI Index 2024, citing Epoch AI estimates: training compute for GPT-4 cost about $78 million and for Gemini Ultra about $191 million at cloud rental pricesBusiness & markets · Large language model · hai.stanford.edu ↗
- Google unveils Gemini (Ultra, Pro, Nano); Gemini Pro starts powering BardSoftware development · Business & markets · Large language model · bostonglobe.com ↗