Top 10 AI Stories Today — 17 August 2026

1. Bank of England Tests Digital Pound and Stablecoin Cross-Border Settlement

FinTech Futures · Finance & Banking

The Bank of England’s Digital Pound Lab has entered Phase 2, selecting Polygon Labs, NOBO Finance, and Dun & Bradstreet to test whether a stablecoin and a simulated digital pound can settle a single cross-border payment together. The experiment also aims to build portable credit identities for small businesses engaged in international trade. While no real customers or money are involved, findings will feed into a Bank of England and Treasury assessment of digital money interoperability due later in 2026.

https://www.fintechfutures.com/fintech-innovation/polygon-labs-joins-bank-of-england-s-digital-pound-lab

2. NYC Bar Calls for Nationwide AI Rules Framework for Lawyers

Law360 / NYC Bar Association · Law & Legal

The New York City Bar Association published a policy paper concluding that AI tools can assist legal work but cannot replace professional legal judgment, and called for a national regulatory framework embedded in the Model Rules of Professional Conduct. The paper highlights an urgent gap: AI is now widely used in drafting, research, and document review, yet no uniform industry-wide rules govern its use by lawyers. Current obligations are spread across a patchwork of ABA guidelines, state bar opinions, and judicial orders, creating uneven standards across jurisdictions.

https://www.nycbar.org/reports/policy-paper-on-the-use-of-ai-tools-by-legal-professionals-in-emerging-companies-and-venture-capital/

3. AI Framework Teaches Biology of Antibodies, Boosting Drug Discovery

News-Medical.net · Healthcare & Pharmaceutical

Boston University researchers published a new antibody-specific AI language model in the Nature Portfolio journal Communications AI & Computing that improved predictions of antibody binding strength by up to 27% while requiring far fewer computational resources than existing models. Rather than treating antibodies like generic proteins, the model was trained specifically on the biological regions responsible for antigen binding, allowing it to better identify promising drug candidates before lab testing begins. Scientists believe the approach could transform therapeutic antibody development, accelerating the path to better medicines, diagnostic tests, and vaccines.

https://www.news-medical.net/news/20260813/New-antibody-specific-AI-framework-accelerates-discovery-of-therapeutic-drugs.aspx

4. OpenAI Activates Advanced Matching for ChatGPT Ads on August 17

Search Engine Roundtable · Media & Marketing

OpenAI today enabled Automatic Advanced Matching (AAM) for all existing ChatGPT Ads Web pixels, upgrading how the platform links conversions back to ad campaigns using hashed customer data from website forms. The update is part of a broader wave of ChatGPT Ads improvements — including a new multi-product carousel format and expansion into the UK, Brazil, and Mexico — that are rapidly maturing the platform into a serious rival to Google and Meta for performance advertisers. With ChatGPT Ads now live in the UK as of August 11, British brands and agencies are being urged to act early while inventory remains relatively uncrowded.

https://www.seroundtable.com/openai-chatgpt-ads-updates-41828.html

5. US Army Deploys Generative AI to Cut Battlefield Planning Time

Federal News Network / OpenPR (DataM Intelligence) · Defense & Security

The U.S. Army is evaluating generative AI decision-support tools that rapidly generate and compare operational plans, reducing the time commanders need to make battlefield decisions. The initiative is part of the broader Next Generation Command and Control programme, which integrates AI with digital communications to improve coordination. It reflects a wider Pentagon push to move AI from experimentation into frontline operational use.

https://www.openpr.com/news/4595831/how-is-military-ai-transforming-modern-warfare-in-2026-latest

6. Aurora Triples Driverless Truck Routes Across Texas as Freight AV Scales

Supply Chain 24/7 · Transportation & Logistics

Aurora has tripled its driverless truck network to 10 routes and added a 1,000-mile Texas-to-Arizona lane as it scales autonomous freight operations. The expansion reflects a broader market shift, with McKinsey reporting over 700,000 fully autonomous robo-taxi rides per week globally. Autonomous trucking is moving from pilot programmes into genuine commercial operations.

https://www.supplychain247.com/topic/tag/Autonomous_Vehicles

7. CrowdStrike Report: AI Now Embedded in Every Stage of Adversary Attacks

CrowdStrike Investor Relations · AI Cybersecurity

CrowdStrike’s 2026 Threat Hunting Report reveals that threat actors are now using AI to generate malware payloads, exploit enterprise AI systems, and scale attacks — with one campaign sending nearly 200,000 AI model requests in just two minutes. North Korea-linked group STARDUST CHOLLIMA injected a malicious package into 131 trusted AI frameworks, while 87% of software registry threats involved malicious npm packages. AI agent-triggered detections are growing at 2.5x the rate of human-triggered ones, putting enormous pressure on enterprise security teams.

https://ir.crowdstrike.com/news-releases/news-release-details/crowdstrike-2026-threat-hunting-report-ai-now-embedded-across

8. Smart Cities Cross Operational Threshold as AI Spending Hits $189 Billion

Solid Market Research · Construction & Infrastructure

Global smart city technology spending reached $189 billion in 2025 — a 22% year-on-year increase — as cities shift from AI pilot projects to full operational deployments, according to IDC. Cities including Orlando are piloting AI for emergency response, climate resilience, and infrastructure stress management, with Europe leading adoption outside China. US programmes under the Bipartisan Infrastructure Law and Inflation Reduction Act have allocated over $65 billion toward grid modernisation and resilient infrastructure.

https://www.solidmarketresearch.com/post/smart-cities-hit-operational-scale-why-2026-is-the-year-pilots-became-platforms