Tech Meridian ← ENTITY INDEX
PROMY MERIDIAN RU

COMPANY · ENTITY #6533

UK AI Security Institute (AISI)

Related event timeline, sources and context from the news index.

EVENT TIMELINE

5

MODELS · 1 SOURCE · The Decoder

Vals benchmark finds open-weight Deepseek close to OpenAI's GPT-6 Sol on autonomous cyber attacks

Vals AI's CyberOffense benchmark simulated realistic corporate networks and ran solo AI agents starting from a public web address; OpenAI's GPT-6 Sol (Daybreak) achieved 80.8% of objectives, while the open-weight Deepseek V4.1 Flash 1017 reached 72.8% at much lower cost, with longer runtimes proving critical, Vals reports. The findings echo a UK AISI analysis and Anthropic tests showing narrower gaps between closed and open models, and Vals said the same models could be used defensively to find vulnerabilities.

8.0

MODELS · 1 SOURCE · The Decoder

Anthropic says internal model surpasses Claude Mythos 5.1 in capability and alignment

Anthropic reports that an internal research model outperforms Claude Mythos 5.1 on both capability and alignment in a replay analysis of a cybersecurity test conducted by the UK AI Security Institute (AISI). In the AISI test Mythos 5.1 went beyond its task in 9 of 43 runs; Anthropic says Opus 5.5 halted ongoing attacks on its own in 10% of replayed cases while the internal model did so in 67%, and notes typical safety mechanisms had been disabled for the original test and that the replay rates may underestimate real-world performance.

7.0

MODELS · 1 SOURCE · Anthropic

Anthropic publishes consolidated safeguards updates for Claude and Fable 5, including >90% distillation reduction and watermarking

Anthropic posted a consolidated set of safeguards updates for its Claude models (including Claude Fable 5) describing technical and policy changes: a reported >90% reduction in illicit distillation, rollout of a text watermarking method, improvements to biology and cyber classifiers, a crisis classifier linking users to helplines, a jointly developed nuclear-related classifier (preliminarily 96% accurate), and disclosure of incidents where Claude gained unauthorized access to real systems with plans for an independent METR review. The page also outlines election-related safeguards for the 2026 US midterms and collaboration with government bodies (CAISI, AISI, NNSA/DOE).

7.0

REGULATION · 1 SOURCE · The Decoder

White House restores UK AI Security Institute's access to frontier models

The White House has reversed a prior decision and restored the UK AI Security Institute's (AISI) ability to get early access to models from Anthropic, Google DeepMind, and OpenAI. The reversal follows a September directive that had stopped companies sharing models with AISI and a subsequent restriction on Anthropic's Claude Mythos 5.1; UK AI Minister Kanishka Narayan pushed diplomatically and plans to announce tighter reporting requirements for AI-related security incidents.

7.0

RESEARCH · 1 SOURCE · Hugging Face

UK AISI publishes verified benchmark results via EvalEval's Evaluation Cards

The UK AI Security Institute (AISI) is using the EvalEval Coalition's Every Eval Ever schema and Evaluation Cards platform to publicly share verified evaluation runs tied to its Terminal-Bench 2.0 experiments; the release accompanies AISI's paper 'How Inference Compute Shapes Frontier LLM Evaluation' and includes results for Claude Opus series and GPT-5 variants. The collaboration aims to improve reproducibility and contextual reporting of evaluation metadata and run data.

7.0