AGP Executive Report

Job Scams: Fake remote and “task-based” job offers are tricking people into paying for training, equipment, or “unlock” fees—legit employers won’t ask for money up front. University ICT Funding: Pakistan’s Sindh University approved a Rs9.44bn 2026-27 budget and backed a new ICT policy to push digital transformation and research. Pixel App Lock: Google is testing an app-by-app lock for Pixel phones on Android 17, potentially using biometrics instead of PINs/passwords. Car Software Updates: Nissan issued a recall for 13,795 X-Trail e-Power SUVs over a lithium-battery controller software issue that can force fail-safe power loss; Honda is also rolling out Gemini-powered conversational AI to select 2026 models. Kenya Presidential Site Hack: Kenya says its president’s website was defaced and taken offline after a cyberattack, but reports no proof of sensitive data compromise. WhatsApp/Banks Attacks: Pakistan’s NCIA warns of rising WhatsApp and bank account hacking and plans a dedicated helpline plus a Meta-backed awareness push. AI Security Risk: Researchers describe “HalluSquatting,” where AI assistants can be manipulated into downloading the wrong software and enabling malware delivery. Closed-Loop BCI Neurofeedback: A new brain-computer interface decodes EEG “error positivity” to train people to detect micro-errors faster, using real-time feedback from brain signals.

EU Regulation Push: The European Commission ordered Google to open up key Android features so users can switch AI assistants beyond Gemini, with changes due by Aug. 1, 2027. Cybersecurity Alerts: CISA warned of actively exploited Microsoft SharePoint flaws and also flagged a Citrix Secure Access Client bug that can let low-privileged Windows users gain SYSTEM access. Platform & App Safety: San Francisco ordered Apple and Google to remove “nudify” AI apps from App Stores, while Meta creators allege a copyright loophole is being used for extortion via fraudulent strikes. Government Digital Services: Pakistan’s Ministry of IT plans a nationwide rollout of standardized Asaan Khidmat Centers for one-window citizen services. AI in Public Sector: Adobe’s Digital Government Index says India is moving toward more integrated, accessible, and AI-ready services. Business Tech: Microsoft is reportedly building “Project Perception” to detect and help fix software vulnerabilities using multiple AI models. Cybercrime & Fraud: Pakistan’s NCCIA says WhatsApp hacking is mainly driven by OTP sharing, and Dubai Police warned about fake visa/job offers spreading via social apps. Education Tech: West Bengal will set up Digital Content Creator Labs in 500 schools to train students in AVGC-XR skills.

Cybersecurity & Courts: A Snyder County jury found a Middleburg man, a cybersecurity company owner, not guilty of misdemeanor stalking and disorderly conduct, after a two-day trial. AI Governance: China released an international AI ethics governance action plan at the World AI Conference in Shanghai, pushing life-cycle ethics, risk tiering, and standards, while also promising open-source high-performance models. Elections Security: Ohio election officials say voter registration and voting machines stay offline to protect personal data, with staff trained against phishing. App Stores Crackdown: San Francisco ordered Apple and Google to purge “nudify” apps tied to non-consensual deepfake porn, citing state criminal and civil rules. Data Center Policy: Apple Valley approved a one-year moratorium on new data center development as other Minnesota cities do similar pauses amid lawsuits. AI Investment Metrics: OpenAI urged companies to measure AI by work completed—successful tasks and cost per success—arguing old software adoption metrics miss AI value. Robotics Safety: Zoox recalled robotaxi software after heavy-smoke detection issues near an active fire scene. Tech Infrastructure: Construction began on the UK’s next national supercomputer at the University of Edinburgh, targeting major research speedups.

Databricks Funding Surge: Coatue-led term sheet values Databricks at $188B, adding to its earlier $5B raise and underscoring investor bets on AI data platforms. AI Skills for Youth: Ucom and Microsoft Innovation Center backed a cybersecurity education program for 14–17-year-olds, handing out completion certificates and focusing on safe online habits. SaaS Security Blind Spot: A new analysis argues security teams can’t rely on infrastructure checks alone; SaaS misconfigurations can expose data quietly. Cybercrime via Fake Repos: Arctic Wolf Labs says attackers used hundreds of impersonation GitHub orgs and repos to spread malware through trusted-looking downloads. Telegram Platform Blocking Debate: India’s Telegram ruling broadens when governments can block entire platforms, shifting scrutiny toward platform design and governance. Real-World AI Breakthrough: Researchers recognized by the Chen Institute and Science Prize use AI speech neuroprostheses to help people rebuild speech from brain signals. Cyberattack Fallout: Japan’s Nichirei restarted parts of its systems after a cyberattack disrupted frozen food logistics. Education + Jobs: SM Supermalls and DICT-CAR plan a free digital upskilling session for jobseekers, including cybersecurity and eGovPH guidance. Fraud Alert: FBI Detroit warns residents that AI-powered scams are spreading across calls, texts, emails, and apps, urging verification before acting.

AI in Android & EU rules: The European Commission ordered Google to let rival AI assistants get the same system-level Android access as Gemini and to share key search data, starting Jan. 2027. Cybersecurity: Fortinet says Jamaica saw 5.4 million cyberattack attempts in the first half of 2026, with AI and automation boosting attackers. AI safety research: Georgia Tech unveiled FIRA, a tool to analyze AI drone crashes and help determine whether model tampering caused failures. Enterprise software & compliance: Regulators moved to cut how much sensitive bank exam data is collected and stored, aiming to reduce breach risk. Markets & chips: Chip sell-offs dragged the Nasdaq and S&P 500, with memory and Nvidia-related weakness weighing on AI trade. Mobile retail IT: Tri Star Energy is consolidating its convenience-store apps onto Rovertown, preserving Paytronix loyalty while centralizing ordering and offers. Education testing: IELTS in India shifts fully to computer-based from September, with paper tests ending in August. AI in banking apps: Visa launched a white-label AI financial assistant for banks, starting US pilots in August.

AI Hardware & Dev Tools: OpenAI launched Codex Micro, a limited-edition desktop keypad that lets developers manage multiple AI coding agents with physical controls and an approval button. Cybersecurity Governance: The White House rolled out the “Gold Eagle” initiative to coordinate AI-discovered software flaw sharing and patching across critical infrastructure. Enterprise AI Oversight: A new push in AI-driven warehouse staffing at Amazon is meeting resistance, with managers overriding software recommendations—highlighting the need for guardrails, not just suggestions. AI + Security Training: Korea’s science minister said the country will train a sovereign AI model focused on cybersecurity by year-end, alongside legal groundwork for ethical hacking. Cloud & Industry Apps: Airbus signed with Scaleway to host industrial and defense apps under EU-focused data sovereignty rules. Semiconductors & AI Supply Chain: TSMC pledged another $100B to expand U.S. chipmaking capacity, lifting total U.S. investment to $265B as AI demand drives record profits. Funding & Startups: India’s Emergent raised $130M Series C to expand an autonomous AI platform for building production-ready apps without coding. Education Tech: Maharashtra postponed MPSC CBT exams to August 2027, keeping offline pen-and-paper until then. Regional Tech Diplomacy: India and the EU agreed to deepen cooperation via an upgraded Trade and Technology Council roadmap covering semiconductors, AI, digital connectivity, and resilient supply chains.

Gold Eagle Cybersecurity Initiative: The White House launched “Gold Eagle,” an AI-powered clearinghouse to detect and prioritize cybersecurity vulnerabilities and coordinate faster patching across critical infrastructure, working with Treasury, DHS/CISA, and the Department of War plus open-source and industry partners. Federal Cyber Spending Push: OMB’s Nick Polk urged agencies to justify cybersecurity budgets with performance and risk-based investment decisions, warning against defending everything at 100% and calling for consolidation of overlapping tools. Defense Tech Procurement: The Department of War’s APFIT program selected 24 FY26 projects, including its first software-only capability awards, as it ramps AI and autonomous systems efforts. Mac Malware Alert: Jamf researchers flagged “CrashStealer,” a macOS info-stealer that impersonates Apple crash reporting and is being actively used by criminals. AI in Education: CU Boulder will require students to complete an AI training module before accessing its ChatGPT Edu service. Meta Face Recognition Debate: Meta removed NameTag code from its Meta AI app after reporting disputes, reigniting questions about whether “inactive” features count as real. Nuclear Plant Data Leak: Reuters reports World Leaks posted thousands of files tied to India’s Kudankulam nuclear plant, raising safety concerns. Android App Store Shakeup: Google and Epic’s settlement collapse means third-party app stores will be allowed on Google Play in the U.S. Visa’s AI for Banks: Visa unveiled an AI Financial Assistant to deliver personalized spending and card insights inside bank apps.

AI Cybersecurity Push: The White House unveiled “Gold Eagle,” a new U.S. clearinghouse meant to centralize AI-discovered software flaws and coordinate faster fixes across federal agencies and critical infrastructure partners. Software Supply-Chain Threats: Checkmarx rolled out self-healing autonomous agents to detect, fix, and verify vulnerabilities during coding, while Arctic Wolf flagged BoryptGrab malware spreading via 292 fake GitHub repositories impersonating real tools. Enterprise AI Tools: Oracle launched an AI-native builder for agentic apps inside Oracle Fusion, aiming to make secure, governed development easier for business and technical teams. Big Tech Shockwave: IBM shares plunged about 25% after an earnings warning tied to customers shifting spending toward AI hardware, wiping out tens of billions in market value and rattling software peers. Real-World Cyber Risk: Reuters reported ransomware group World Leaks posting alleged files tied to India’s Kudankulam nuclear plant, highlighting how third-party data centers can become weak links. Local Tech & Services: Bosnia’s Kliktura expanded globally with bilingual web/SEO/AI automation offerings, and South Carolina’s Tax Free Weekend (Aug 7–9) includes computers and tech.

AI & Security Policy: The U.S. will create a coordination group to share AI-identified cybersecurity vulnerabilities and coordinate responses across software, AI, and critical infrastructure providers. Cybercrime & Critical Infrastructure: U.S. authorities unsealed charges and offered a $10M reward tied to Russians accused of ransomware-related infrastructure work targeting victims across 21 states. Enterprise Cybersecurity Shift: ISC2 says cybersecurity pros are spending more time validating AI recommendations, with many reporting AI mistakes and humans still held accountable. Corporate Tech Pressure: IBM shares plunged after preliminary Q2 results missed expectations, as clients reportedly shifted capex toward supply-constrained AI infrastructure while cybersecurity concerns stayed top of mind. Privacy & Compliance: California’s ALPR law is driving new written policy requirements for plate-reading systems, with $2,500 minimum damages per violation. Healthcare Risk Gap: A Fortified Health Security report finds healthcare visibility into cybersecurity risks is improving, but remediation remains slow. Retail Payments Modernization: GK Software USA and National Retail Solutions push integrated mobile and POS/payment ecosystems to help grocers and neighborhood retailers speed checkout and manage multiple processors. Tech in Vehicles: Tesla Korea is moving FSD to a lower-cost monthly subscription, putting pressure on Hyundai’s software strategy. Mobile Device Support: Xiaomi has ended software support for multiple phone and tablet models, signaling no further updates or security fixes.

Software Supply Chain Security: JFrog’s 2026 “State of the Union” report warns software supply chain risk is surging as attackers shift from package registries to AI model registries and development tools, exposing a governance blind spot. Cyber Threats & Critical Infrastructure: CISA says Russian FSB-linked actors are exploiting misconfigured routers to probe and attack critical sectors, while Resecurity flags rising reconnaissance aimed at mapping exposed AI systems back to organizations. AI Governance & Deepfakes: KnowBe4 reports “shadow AI” and deepfakes are outpacing controls, with many employees admitting they could be tricked by realistic scams. Quantum Computing Push: Quantinuum, Rolls-Royce, Riverlane and EPCC sign an agreement to explore fault-tolerant quantum computing for industrial simulations like gas turbine design. Productivity & Platforms: OpenAI launches “ChatGPT Work” to automate app and task workflows, and Apple rolls out iOS 27 public beta with a revamped Siri AI. Enterprise Cloud: AWS marks 20 years of SQS, highlighting how message queuing helps isolate failures and handle spikes. Real-World Tech in Healthcare: Olio will be deployed by Cantex to coordinate home health and SNF referrals and improve visibility across care transitions.

Defense Cyber Rules: The Pentagon is pausing CMMC Phase 2, keeping only Level 1–2 self-assessments while it reviews how to cut compliance burdens for small defense suppliers. Cybersecurity Response: CISA published lessons from a GitHub credential exposure incident, detailing how it responded and what internal gaps it’s fixing. Ransomware Sanctions: OFAC sanctioned VPN and “cryptor” providers tied to ransomware operations targeting Americans, coordinating with the UK. AI in Public Services: Dane County, Wisconsin will deploy an AI assistant to handle non-emergency calls in multiple languages, escalating to live dispatchers when needed. Local Breach Notice: St. Paul, Minnesota notified residents after a last-summer cyberattack that exposed personal files from a shared drive. Enterprise Software: Accela bought Civira AI to automate public-sector permitting and licensing workflows. Consumer Tech: Apple opened the first public beta of iOS 27, featuring a revamped Siri AI. Policy & Privacy: India is weighing uniform username rules across messaging apps, raising new questions for end-to-end encryption.

AI & Cybersecurity: MeitY Secretary S Krishnan says AI can help defenders act faster, as India launches the Digital Threat Report 2025-26 and stresses faster threat exploitation. Messaging Rules: India is reviewing a proposed username feature across messaging apps after WhatsApp paused rollout over fraud and impersonation risks. Fraud via Social Apps: Uzbekistan court documents describe Telegram account takeovers and fake law-enforcement identities used to steal tens of millions of soums. Enterprise AI Hiring: TCS plans to deploy up to 8,900 forward-deployed AI engineers and is also expanding AI and cybersecurity services for clients. Cloud/AI Integration: LTTS shares rose after an Anthropic Claude integration into its BlueVerse platform, including a Claude Center of Excellence. Hardware Costs: An AI-driven memory shortage is pushing up prices for iPads, consoles and laptops, with shortages expected to last into 2028. Gov IT Systems: Nepal’s Department of Passports went live on a new system after migrating ~10 million records. Policy & Safety: Meta’s CSAM ads response is under examination by India’s IT ministry. AI Debate: A prominent scientist argues human-level AI may be out of reach, challenging assumptions behind the Turing test and modern AI research.

AI & Society: Hundreds protested at OpenAI, Anthropic, and DeepMind sites in San Francisco, ending with pizza and dancing—another sign AI is becoming a mainstream political flashpoint. Education Tech & Policy: Illinois signed a law banning AI in teacher evaluations, while U.S. schools struggle to block kids from risky games and content online. Cybersecurity Regulation: Australia’s push to upgrade the ACSC into a formal cybersecurity regulator aims to move from voluntary guidance to enforceable standards and incident powers. Enterprise AI Security: IBM and Red Hat launched Lightwell to automate open-source vulnerability remediation with certified fixes. Digital Transformation Partnerships: Kuwait’s CITRA and Huawei Kuwait signed an MoU to accelerate AI, cloud, and 5G initiatives under Kuwait Vision 2035. Crypto Basics: Multiple explainers hit the same point: crypto wallets don’t “store” coins—keys do. Local Tech in Services: Egypt’s Grand Egyptian Museum expanded official ticketing via telecom apps to reduce fake links and improve secure access. Workforce & Defense: A new U.S. defense/aerospace white paper flags workforce, infrastructure, advanced manufacturing, and innovation as the 2026 priorities.

AI & Cybersecurity Risk: JPMorgan CEO Jamie Dimon says Anthropic’s Mythos AI model found thousands of unpatched software flaws, warning organizations have a limited window to fix them before exploitation risk rises. Enterprise IT & Growth: TCS reported 13.93% YoY revenue growth in Q1 FY27 (₹72,275 crore) and outlined continued AI investment plus a shift toward acquisitions and training for AI, data security and cybersecurity. Digital Payments: RCBC will make unlimited free InstaPay and PESONet P2P transfers via its Pulz and DiskarTech apps, starting July 15, cutting everyday transfer fees. Govt Tech & Skills: Qatar launched “Maharat 2026” summer training for students across sectors including IT and digital services, while KT created an information protection advisory committee covering AI security, zero-trust, cloud security and response. Security & Fraud Awareness: A surge in fake antivirus renewal pop-ups is pushing users to pay and hand over card details via browser notifications. Tech Policy & Content Controls: India’s inter-department committee says the ban on the film “Satluj” under the IT Act is justified, keeping it off OTT platforms while review continues.

AI Policy & Governance: India’s MeitY is reportedly drafting uniform rules for messaging apps after WhatsApp’s username feature sparked fears of phishing, impersonation, fraud and “digital arrest” scams. Cybersecurity Awareness: Dubai Police warned residents not to download untrusted apps, stressing that fake installers can steal banking data and funds. Education Tech: dti Publishing’s NetEmulator® 3.0 (Cybersecurity Edition) adds hands-on firewall, secure access control, attack-surface reduction, vulnerability testing and intrusion detection training. Fraud & Incident Response: Sri Lanka’s SLCERT says indirect online fraud complaints have risen, urging users to watch for suspicious links and requests for personal info. Enterprise/Health IT: Olio will be deployed by Cantex to coordinate home health and SNF referrals and improve visibility across value-based care transitions. Hardware Reliability: Philips will replace affected Hue Bridge Pro units bricked by a specific June firmware update scenario. Infrastructure & Connectivity: Angola’s first Technology and Science Fair used Angosat-2 to provide temporary internet access for students and visitors. Security in the Wild: CISA says it built its incident playbook during an active breach, highlighting gaps in preparedness.

Messaging Regulation Shake-Up (India): India’s IT ministry is weighing uniform rules for messaging apps after WhatsApp’s proposed username feature sparked fears of fraud, impersonation, and phishing—MeitY plans consultations with major platforms before finalizing a common framework. Cybersecurity Readiness Gap (US): CISA admitted it lacked a pre-made incident response playbook when a breach hit, saying staff had to build one during the early stages after a contractor’s exposed credentials were found on a public GitHub repo. AI Model Access & Privacy (Meta): Meta released Muse Spark 1.1 with a paid API for agentic coding and tool use, while Muse Image drew privacy backlash for generating images from public Instagram content by handle reference unless users opt out. Windows Tracking Concern (Microsoft): Microsoft acknowledged Windows uses a persistent Global Device Identifier that can’t be turned off, limiting user control over cross-service activity linkage. AI Security Warning (China): China’s NVDB warned that versions of Anthropic’s Claude Code may contain a back door, urging uninstall/upgrade amid escalating scrutiny of foreign AI tools. Digital Infrastructure Talent (US): Data Center Youngbloods expanded its global talent pipeline to connect early-career workers with data center roles via mentorship and operator partnerships. OTT Film Certification Fight (India): The Satluj controversy continues as a government committee reportedly recommends keeping restrictions, with renewed debate over whether OTT releases should require CBFC certification.