AI
Homebase Assistant automates shift cover and payroll checks for small teams
New AI tool handles scheduling gaps and timecard errors while keeping owners in control of final decisions.
AI
AI coverage from Vector Bulletin holds 67 articles, 17 of them reference guides. The newest was published on October 20, 2026. New stories are added as soon as they are confirmed, from more than 150 sources checked as often as every 45 seconds. Each story lists its sources. Primary sources we follow for this section include MITRE ATLAS and NIST AI Risk Management Framework.
AI
Microsoft 365 E7 Guardian Agents Apply Dynamic Permissions to Mitigate AI Insider Threats
New tools treat autonomous AI agents as potential insider threats, using continuous risk assessments to manage permissions and access controls.
AI
AI Influencer Scam: Man Travels 2000km to Meet Fictional Persona
A Brazilian man sold his possessions to visit an AI-generated woman, highlighting the risks of parasocial relationships with synthetic media.
AI
UK financial regulator demands AI failure and near-miss reporting from firms
The FCA states that incident reporting requirements must extend to significant AI failures and near-misses to ensure accountability for autonomous systems.
AI
Former OpenAI Engineer Detained for Alleged Executive Room Surveillance
A US court ordered the detention of a former staff member accused of planting a hidden recorder in a San Francisco office to capture sensitive conversations.
AI
doubleAI Raises $200M for Binary Optimisation Model Maxwell
Tel Aviv-based doubleAI secured a $200 million Series A round, with Nvidia and Dell joining Lightspeed and Bessemer in the investment.
AI
Anthropic-Nscale $45B Talks Face Hurdle as Investors Challenge $30B Valuation
Bloomberg reports that Anthropic and Nscale are discussing another massive infrastructure deal, mirroring their previous agreement announced in August.
AI
New York Mayor Mamdani urges resistance to AI inevitability narrative
Zohran Mamdani argues that public officials must regulate artificial intelligence rather than accept unchecked adoption by private companies.
AI
Cloudflare Clairic: Structured Data Extraction Cuts LLM Costs by 49% and Latency
Cloudflare’s preview model extracts structured data from text and images in one pass, avoiding token costs and reducing latency compared to large language models.
AI
OpenAI ChatGPT Research Highlights Improved Coding Support for COBOL and SQLite Migration Tasks
OpenAI releases research detailing improvements to ChatGPT's coding capabilities for less common languages, including new evaluations for COBOL and SQLite.
AI
Microsoft introduces Guardian Agents to monitor internal AI behaviour
Microsoft proposes a new security architecture to supervise autonomous AI agents operating within enterprise networks.
AI
OpenAI Engineer Detained After Alleged Conference Room Surveillance
A federal judge denied bail for a former OpenAI engineer accused of bugging a meeting space, citing concerns over foreign government ties.
AI
STAR Fuzzer AI Agents Write Harnesses, Finding Critical Bugs in 45% of Projects Without Running Apps
An agentic fuzzer writes its own test harnesses to find crashes in critical code without needing a running application.
AI
Google Gemini Adds Virtual Try-On and Garment Adjustment Features
Athlete stylist Sydney Bordonaro uses Gemini to visualise outfits remotely, swap items and get cleaning tips for stains on professional wear.
AI
Russia’s AI-driven VPN blocking plan risks mass false positives
Draft national strategy proposes using artificial intelligence to identify circumvention tools, raising concerns over collateral damage to legitimate traffic.
AI
AI Security Engineers See 268% Surge as US Cyber Postings Drop 26%
CyberSN data shows overall US hiring fell 26% in the first nine months of 2026, while AI security roles saw a 268% increase compared to the same period last year.
AI
AI transparency definition and how it works in practice
AI transparency reveals how software models make decisions, yet complete visibility often exposes proprietary secrets to competitors.
AI
Prompt injection recovery in AI desktop apps: containment steps
Desktop AI applications execute local code through hidden instruction channels that require strict process isolation to reverse.
AI
AI coding assistant supply-chain risks: spotting poisoned code generation
Detecting malicious code injected via software development utilities requires monitoring build outputs rather than examining user prompts.
AI
EU Updates Research Guidelines to Clarify Accountability for Agentic AI Systems
The European Commission’s latest guidance confirms that researchers remain fully responsible for outputs generated by autonomous AI agents in academic work.
AI
OpenAI Office Found With Surveillance Device in Weekly Report
Neowin reports that an OpenAI office was found to be bugged, highlighting a physical security incident in its weekly technology roundup.
AI
Gemini 4 Argon Agents Autonomously Code 2,783 Programs to Slay Minecraft Boss in 27 Hours
Four AI agents using Google’s Gemini 4 Argon model defeated Minecraft without human input, coordinating tasks and writing thousands of programs to slay the boss.
AI
AI-Modified Photos Distort Human Recollection With Unwarranted Certainty, MIT Study Warns
MIT researchers demonstrate that AI-altered photographs can distort human recollection, leading subjects to hold incorrect memories with unwarranted certainty.
AI
Intel D-series AI accelerator hits 1.3 PFLOPS FP8, beating Nvidia B300 memory capacity
Intel unveils a data centre accelerator with record memory capacity, prioritising volume over traditional graphics performance metrics.