Artificial intelligence
AI for import documents: what Azerbaijan should measure before deployment
Wonderful demonstrated import-document checks. Evaluating a pilot requires time to acceptance, error rates and cost per completed application.
LG CNS completes a 30 MW AI data center in Jakarta
A customer committed to all first-phase capacity before construction finished. Expansion to 400 MW remains a plan, not current capacity.
Marvell agrees Lumentum combination in $115 billion deal
The all-stock deal would combine Marvell chips with Lumentum optical technology. Shareholder and regulatory approvals are still required.
Prompting as a specification: a framework for more controlled AI workflows
A paper in Acta Innovations sets out the CORE framework for AI prompts. FinFly explains its structure, a newsroom example and the limits of the approach.
Marble raises $3.75 million for AI that learns workplace processes
The Munich startup studies how teams work before proposing automation. Heartcore and Seedcamp co-led the company's pre-seed funding round.
Anthropic plans $3.5 billion share sale to Japanese investors in IPO
The proposed sale forms part of a US IPO. The Japanese tranche does not establish Anthropic's total valuation or the size of the full offering.
ADNOC partners with OpenAI and G42 to bring AI into energy operations
The partnership targets operational and commercial workflows, with human oversight retained for consequential decisions across the business.
Samsung’s chip profits contrast with losses in its device division
Samsung’s results show different economics for component suppliers and finished devices. AI-related demand does not benefit every segment equally.
AWS growth comes alongside negative free cash flow at Amazon
AWS is growing rapidly, but Amazon’s infrastructure spending exceeds operating cash flow. Profitability and cash generation answer different questions.
Serval raises $150m as enterprise AI puts implementation economics in focus
Serval’s $3bn valuation and planned automation-engineer hiring highlight the work required to turn enterprise AI into functioning processes.
Poolside to shut down its AI chat service on 31 October
Poolside Chat will shut down on 31 October. Users are being directed to downloadable models that can run on their own infrastructure.
Polygraf AI wins award for enterprise AI usage controls
Polygraf AI is among the 2026 Top InfoSec Innovators. Its technology checks exchanges with AI tools and applies organizational security policies.
Perplexity updates AI models for data classification and request routing
The Decisions API classifies requests and scores data. Perplexity’s October update introduces v1.2 models and enterprise connector controls.
ASUS shows ProArt GR1X mini PC for always-on AI workloads
The ProArt GR1X fits into a 150 × 150 × 51 mm chassis. ASUS targets local background AI workloads but does not state a price in the announcement.
Azerbaijan's Finance Ministry to Use AI for Public Finance Analysis
Finance Minister Sahil Babayev said artificial intelligence will be used to analyze and forecast public finances and automatically detect anomalies.
ABB Bank Hosts International Risk Management Forum in Baku
The International Risk Forum, initiated by ABB Bank and supported by Mastercard, brought together regulators, bankers and experts in Baku to discuss geopolitical uncertainty, AI, cyber threats and digital assets.
OpenAI expects to reach $70B in annualized revenue by year-end
AI model developer OpenAI expects annualized revenue to reach or exceed $70B by the end of the year. At the end of September, the figure was estimated at about $50B.
New York City Comptroller Urges Anthropic to Delay IPO Over AI Risks
Mark Levine has asked Anthropic CEO Dario Amodei to hold off on the company's stock market listing until it addresses safety, development pace and corporate governance risks.
Share of AI-Generated Articles in Russian Media Reaches 18%
By the end of Q2 2026, 18% of materials in Russian media show signs of neural network generation. The study also found that direct quotes and expert comments are disappearing from texts alongside AI.
Google Opens Public Access to Gemini 4 Argon, Its First Fourth-Generation Model
On October 13, 2026, Google began general rollout of Gemini 4 Argon. The model is now available to developers, companies, scientists, and Google AI Pro and Ultra subscribers after a two-week closed testing period.
Azerbaijan Rolls Out AI in Agriculture: 13 Solutions Already Operational
Under the Agrarian Artificial Intelligence project, 13 solutions have been deployed and 11 more are being tested. Deputy Minister of Agriculture Ilhama Gadimova provided an update on the project's progress.
Azerbaijani Farmers Gain Access to 'Agrarian AI' Services via E-Cabinet
Around 350,000 active farmers in Azerbaijan will be able to use tools from the 'Agrarian Artificial Intelligence' project through their personal e-cabinets, Deputy Minister of Agriculture Ilhama Gadimova said.
AWS Agentic Football Cup with AI-Controlled Teams to Be Held in Baku
On October 17, Baku will host the AWS Agentic Football Cup, organized by AWS User Group Baku with support from Azercell. Participants will create virtual football teams controlled by AI agents.
AI Errors Emerge as a New Risk for Auditing and Business
A study by international authors has found that commercial language models can exhibit "functional deception," creating new operational risks for companies and prompting a rethink of AI auditing and control.