OpenAI launches ChatGPT “Study Mode” to promote responsible learning 

OpenAI has introduced a “study mode” in ChatGPT, designed to foster responsible academic use by guiding students through topics step-by-step instead of simply supplying answers. Accessible via the tools menu, it adapts explanations according to a learner’s math proficiency and intent. This launch comes amid a steep rise in AI-related cheating in UK universities, with nearly 7,000 confirmed cases in 2023–24. The new mode supports image inputs, developed with input from educators and experts, though OpenAI cautions about inconsistent outputs.

Gemini adds Temporary Chats and new personalization features

Google has updated its Gemini chatbot with two major features, Personal Context and Temporary Chats, available to all users at no cost. Personal Context lets Gemini remember details from past conversations to deliver more relevant, tailored responses. Temporary Chats, by contrast, are one-off sessions that vanish after closing, leaving no trace in history or personalization data. These tools mirror popular ChatGPT features, offering both improved personalization and privacy controls. The rollout begins with Gemini 2.5 Pro in select regions, expanding to Gemini 2.5 Flash later this year, underscoring Google’s push to make AI assistance smarter and more user-friendly.

NetSuite 2025.2 elevates supply chain efficiency with smarter planning tools

NetSuite’s 2025.2 release introduces enhancements to its Materials Requirements Planning (MRP) module, streamlining supply chain planning and execution. The new Planning Workbench Views, Reschedule In Firm, Reschedule Out Firm, and Cancel Firm, offer actionable suggestions when disruptions occur, allowing planners to reschedule or cancel orders more effectively. The Plan Definition Saved Search now supports automated email alerts for failed MRP runs, saving planners from manual monitoring. Additionally, VIP orders can be flagged via the new expedite marker on the Planned Orders tab, and transfer orders are now filterable by sufficient stock, saving time on order fulfillment.

SAP evolves S/4HANA EHS from reactive compliance to AI-driven proactivity

SAP’s 2025 strategy update transforms its S/4HANA Environment, Health, and Safety (EHS) Management from a reactive, monolithic system into a modular, connected suite that anticipates compliance needs. Features like intelligent permit management use automated regulatory extraction, while proactive regulatory intelligence prepares organizations for shifting requirements. Workplace safety evolves with AI-assisted instruction generation, integrated risk assessments, and performance analytics, cutting manual workloads by up to 70% and reducing near-misses. SAP envisions the suite advancing toward predictive capabilities, empowering customers with greater operational efficiency and regulatory agility.

PoisonSeed phishing kit bypasses MFA with Adversary-in-the-Middle tactics

A new phishing toolkit named PoisonSeed is actively undermining multi-factor authentication (MFA) through sophisticated adversary-in-the-middle (AitM) methods. Victims receive phishing emails posing as trusted platforms like Google, SendGrid, or Mailchimp. Links embed the target’s encrypted email in the URL and store it as a cookie, verified server-side before advancing to the fake login prompts, this "precision-validated phishing" ensures only intended users proceed. The kit supports multiple 2FA channels (authenticator apps, SMS, email codes, API keys), relaying credentials and capturing authentication cookies to gain full account access.

Google sales CRM breach exposed Ads leads

Google has confirmed that a compromised Salesforce Customer Relationship Management (CRM) instance exposed information on prospective Google Ads clients. The leaked data included basic business contact details like company names, phone numbers, and related notes used by sales agents, but did not include payment info or data from platforms like Google Ads, Merchant Center, or Analytics. The incident, attributed to the cybercriminal group ShinyHunters (also known as Scattered Spider or “Sp1d3rHunters”), reportedly affected around 2.55 million records. Google detected the breach in June, contained it swiftly, enacted mitigations, and has notified impacted parties.