GAI Press.

United Airlines bans AI agents from booking flights, citing cyber risks

Original headline: United Airlines blocks AI agents from booking flights, citing cybersecurity risks

United Airlines announced that it will prohibit customers from using AI agents to book flights or manage reservations. The airline said the measure is intended to mitigate cybersecurity threats.

Restricting AI‑driven booking services signals heightened security concerns and may reshape airline customer‑service strategies.

Research angleHow does the use of AI agents affect airlines' perception of cyber threats? — Protection Motivation Theory

Apache Patches Critical HTTP/2 RCE Vulnerability Discovered by Anthropic and Google

Original headline: Apache Patches Critical HTTP Server RCE Found by Anthropic, Google

Apache announced that it has patched a race‑condition vulnerability (CVE‑2026‑119618) in its mod_http2 module, which could allow unauthenticated attackers to execute code on servers with HTTP/2 enabled. The flaw was jointly discovered by Anthropic and Google, and the update significantly reduces the risk of server compromise.

It lowers the risk of large‑scale remote‑code‑execution attacks on HTTP/2‑enabled web servers.

Research angleHow does the speed of vulnerability patch deployment affect the frequency of security incidents in organizations? — Protection Motivation Theory

Daylight Secures $80 Million to Boost AI‑Powered Security Operations

Original headline: Daylight Lands $80 Million for AI-Powered Security Operations

Daylight announced that it has raised $80 million to enhance its AI‑powered security operations. The funding will be used to accelerate product development, recruit new talent, expand its presence in the United States, and enter additional markets.

The influx of capital underscores growing investment in AI security solutions, intensifying market competition.

Research angleHow does the adoption of AI‑based security operations by firms affect customer trust? — Technology Acceptance Model (TAM)

Nvidia CEO Jensen Huang resigns, raising doubts over AI chip monopoly

Original headline: Jensen Huang built an AI kingdom. Now he’s gone, will Nvidia’s ramparts be breached?

Nvidia CEO Jensen Huang announced his resignation amid a personal scandal, bringing the company's opaque contracts and supply chain into focus. Nvidia's high‑performance GPUs have acted as the AI economy's "central bank," and his departure has sparked concerns about the impact on the foundational infrastructure of the AI sector.

Nvidia's dominance in AI GPUs means leadership changes could destabilize the entire AI ecosystem.

Research angleHow does a change in Nvidia's top leadership affect AI firms' intention to adopt its GPUs? — Technology Acceptance Model (TAM)

Japan Announces Emergency Cybersecurity Funding as AI-Assisted Breaches Surge

Original headline: Japan Pushes Emergency Funding for Cybersecurity as AI-Assisted Breaches Rise

Japan’s government said it will allocate emergency funding to strengthen cybersecurity in response to a rise in AI‑assisted breaches. The package creates public‑private cyber response teams, establishes an AI‑focused unit within the Ministry of Economy, Trade and Industry, and expands the government’s new active cyber‑defense authority.

Boosting funding and authority now equips Japan to rapidly improve its national cyber‑defense amid growing AI‑driven attacks.

Research angleHow does the increase in AI‑facilitated breaches affect trust in public‑private cyber collaboration? — Trust Theory

Google rolls out Gemini 4 Hydrogen, finally replacing 3.8 Flash

Gemini 4 Hydrogen isn't getting a slow, security-conscious release like Carbon and Argon.

Security Researchers Find Fake AI Domains Distributing ClickFix Lures, Browser Hijacking Malware

Original headline: 186 Fake AI Domains Redirect Users to ClickFix Lures, Browser Malware, and Search Hijackers

Security researchers have discovered that look-alike domains masquerading as AI tools are being used to deliver ClickFix-style lures, browser-hijacking malware, and extensions that hijack user searches. This delivery pattern has been dubbed “TypoPrompting” by the research group Island.

The technique creates a new infection vector that traps users when they try to access legitimate AI services.

Research angleHow do fake AI tool domains affect users' security awareness and protective behaviors? — Protection Motivation Theory

Hackers Used AI-Generated Scripts to Breach US Retailer's Network

Original headline: Inside an AI Attack Against a US Retailer

When manual commands failed to penetrate a US retailer’s systems, the attacker resorted to AI-generated scripts. Within minutes the scripts enabled lateral movement across the network, demonstrating rapid compromise using generative AI.

The incident shows that AI-generated code can bypass traditional defenses and accelerate network spread.

Research angleHow does the use of AI-generated attack scripts affect employees’ security awareness? — Protection Motivation Theory

Symantec’s Dick O’Brien Urges Shift from Patching to Mitigation in AI Era

Original headline: Security Teams Must Shift from Patching to Mitigation in the AI Era

Dick O’Brien of Symantec warned that the window for applying patches is closing rapidly as AI technologies evolve. He said organizations can no longer rely solely on patching and should adopt an ‘assume‑exploit’ mindset, focusing on proactive mitigation measures.

Relying only on patches is insufficient against AI‑driven vulnerabilities, making a mitigation‑first strategy essential.

Research angleHow does an assume‑exploit mindset affect organizations’ effectiveness in countering AI‑driven threats? — Protection Motivation Theory