Can AI coding agents rebuild biological AI tools from a paper?
UK AI Security Institute benchmarked coding agents to assess if they can recreate restricted biological AI models from research papers.
UK AI Security Institute·26 Oct 2026
Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
UK AI Security Institute benchmarked coding agents to assess if they can recreate restricted biological AI models from research papers.
Block introduced Dessa, a voice AI model trained on human conversation, which drew both industry praise and concerns over potential scams.
OpenAI announced new security protocols for training AI models in closed systems without internet access.
Anthropic unified Claude chat and Cowork across web, desktop, and mobile platforms for enterprise customers.
An Apollo economist states AI investments have generated zero financial payoff for non-tech corporate America so far.
MetLife Investment Management's Drew Matus discusses Big Tech AI spending risks and market expectations on Bloomberg This Weekend.
Mistral is utilizing open-source models to offer enterprise clients lower costs, data control, and European AI sovereignty.
JPMorgan Private Bank is using AI for client onboarding and investment research to boost adviser productivity in EMEA.
Paris Marx highlights global lessons from data center developments in Ireland and New Zealand for US infrastructure projects.
Apollo Chief Economist Torsten Slok states AI investments have provided zero financial payoff to S&P 500 companies outside tech.
SoftBank plans $50 billion in technology investments in Malaysia to accelerate its global bet on artificial intelligence infrastructure.
Financial Times analysis suggests tech billionaires use philanthropy to secure regulatory goodwill for data centres and AI bot development.
Pennsylvania Governor Josh Shapiro signed a law requiring disclosures on AI-generated advertisements across all media formats.
Block announced a new voice model capable of running locally on consumer hardware like a MacBook M4 Max.
Security researchers used AI tools to uncover legacy vulnerabilities in PHP, prompting subsequent AI-assisted exploitation attempts.
Gibraltar and ADI Foundation signed an MOU to explore a multi-currency stablecoin consortium for institutional cross-border settlement.
Customers Bank reduced select commercial loan closing times from 30-60 days to 7 days in an AI pilot reported in Q2 earnings.
OpenAI announced it will stop training frontier models on the live internet, citing significant safety and integrity risks.
The UK pledged £100 million for an AI Biosecurity Lab, backed by OpenAI, Anthropic, and Google DeepMind, to study frontier risks.
The UK government established a dedicated AI lab to monitor biosecurity threats and risks of weapon creation.
Robinhood CIO Stephanie Guild discussed consumer trading trends and retail investors using agentic AI to invest on their behalf.
AlixPartners Co-CEO Rob Hornby notes AI costs are falling faster than historical technologies, though enterprise returns remain uneven.
OpenAI moved frontier model training to air-gapped environments using cached web content to mitigate cybersecurity risks.
OpenAI has started severing internet access during training for its most advanced models following past safety breaches into other systems.
Databricks highlights growing AI agent usage in writing Spark Declarative Pipelines, promoting declarative data engineering.
Cloudflare updated its Pingora cache architecture to automatically coordinate partial fetches, cutting R2 p99 latency in half.
Microsoft Threat Intelligence detailed a supply chain compromise in LiteLLM version 1.107.0 involving credential theft via Python hooks.
SpaceX launched a partner network with 130 technology leaders to support business deployment of its Grok AI model.
OpenAI reportedly eyes a valuation near $1.25 trillion amid ongoing fundraising talks, while Anthropic approaches a potential IPO.
Anthropic and Hg have reportedly agreed to create an independent consultancy to embed AI engineers in European enterprises by year-end.