AI Safety & Alignment

Today's Stories

1. 1

   ServiceNow CFO: Security Tops AI Barriers, AI Control Tower Eyed

   ServiceNow President and CFO Gina Mastantuono said at Citi's TMT conference that customers cite security and risk management as the biggest obstacles to large-scale AI deployments, and pointed to AI Control Tower as a tool to secure and govern AI agents. ServiceNow says AI usage rose ninefold from Q1 to Q2, more than 50 customers now pay over $1 million for new AI packages, and it raised its 2026 AI target from $1 billion to $1.5 billion.

   Whether AI Control Tower's cost and ROI controls translate into wider deployments, given reported savings such as a 65% reduction in IT service-desk costs for Raleigh and potential savings exceeding $5 million for a European energy company.

2. 2

   Palo Alto CEO Nikesh Arora: AI threats push vendor consolidation

   Palo Alto Networks CEO Nikesh Arora said AI vulnerability-finding models have driven talks with roughly 2,000 companies, and its own first test surfaced about 1,200 potential vulnerabilities. Arora said about 60% of vulnerabilities found in testing came through Mythos, about 30% through OpenAI models, with the rest from others — evidence he says supports a multi-model approach.

   He said consolidation favors incumbents with integrated platforms and may lead customers to cut vendors over time, with changes tied to contract renewals. Watch Palo Alto's $100 million in Prisma AIRS revenue.

3. 3

   AI opens supply chains to hackers, and fights them

   Uber Freight, Ceva Logistics, and Coca-Cola's Fairlife suffered cyber incidents, as AI-powered trackers, cameras, and sensors make warehouses and plants more vulnerable to hacks. A breach can suspend operations for an average of 247 days, per IBM, delaying production and potentially causing products to expire across the supply chain.

   The response hinges on whether companies deploy AI to patch software flaws faster than attackers can exploit them, while also training workers to spot deepfake phishing.

4. 4

   Doctor scolds patient who came after ChatGPT consult

   An office worker who felt chest tightness asked ChatGPT about his symptoms, was told heart or lung problems could not be ruled out, and went to a hospital. The doctor sighed and asked why he came just because AI said so. The tests found no serious abnormality and the doctor said there was little to worry about, yet the visit still drew a rebuke. The poster said he never used AI to overrule the doctor's diagnosis or misuse medicine.

   The dispute hinges on whether AI symptom-checking should count as a reason to seek care, or as needless anxiety. Commenters split, with some saying many patients blindly trust AI and others saying AI cannot make a medical judgment.

5. 5

   NTTデータ先端技術, YCP and ABP recheck 7 SI hypotheses

   NTT Data IntelliLink, YCP and AI Brain Partners jointly published a white paper re-verifying seven March 2026 hypotheses about generative-AI-era systems integration, judging H1 and H3 as having progressed greatly while H2 and H7 were unevaluable. The three firms say SI's competitive axis is shifting from how fast you can build to how safely you can operate, translate into industry context and train successors, with each hypothesis rated on a five-point scale and sources cited.

   They plan monthly editions starting with September 2026, then October and November, updating hypotheses over time around four strategic showdowns such as safety versus results and sovereign AI versus global.

6. 6

   Finnish MP's question draws government reply on superintelligent AI

   A citizen cold-emailed a Finnish MP, met him for lunch, and the MP submitted a written question four days later, drawing a government reply after four weeks. Finland's government said it seeks to "constructively promote the creation of international regulation on the development of superintelligent AI," a formal answer the citizen says stemmed from a cold email.

   The reply paired that commitment with an emphasis on balancing safety with innovation, so the test is whether Finland pushes actual restrictions. Watch the 21-day response rule that forced the government to answer.