ChatGPT’s new memory feature remembers everything you’ve told It 

OpenAI has significantly enhanced ChatGPT's memory capabilities, enabling the AI to recall users' past interactions to deliver more personalized responses. This upgrade allows ChatGPT to consider user preferences and interests in tasks such as writing, learning, and advice. Previously, users could manually manage memory in the "Manage memories" tab, but now ChatGPT can autonomously remember and incorporate details from conversations. Users retain control over their data, with options to disable memory through settings or use temporary chats that do not influence stored memories. The feature is currently rolling out to ChatGPT Plus and Pro users globally, excluding users in the UK, Switzerland, Norway, Iceland, and the European Economic Area (EEA).

AI chatbots face rising threat from 'Zero-Knowledge' hackers

AI chatbots are increasingly vulnerable to exploitation by "zero-knowledge" hackers—individuals with minimal technical expertise who leverage generative AI tools to craft sophisticated cyberattacks. A recent report by Cato Networks highlights how these attackers use techniques like the "Immersive World" jailbreak to bypass safety measures in AI models such as ChatGPT, Copilot, and DeepSeek. Etay Maor, Cato's Chief Security Strategist, emphasized that these guardrails "should have been blocked by gen AI guardrails. It wasn't," underscoring the ease with which AI can be weaponized. The report notes that attackers can trick AI systems into generating malicious code, posing significant risks to organizations. Experts warn that without robust safeguards, the democratization of AI tools could lead to a surge in cyber threats.

NetSuite unveils AI-driven enhancements to empower southeast asian businesses

At SuiteConnect Singapore on April 8, 2025, Oracle NetSuite introduced a suite of AI-driven enhancements aimed at empowering organizations across Southeast Asia. These updates include advanced predictive, generative, and agentic AI capabilities designed to streamline operations, enhance productivity, and provide deeper business insights. Key features encompass NetSuite Text Enhance, Prompt Studio, and a Financial Exception Management agent, all integrated to improve decision-making and reduce manual errors. Additionally, the launch of NetSuite Enterprise Performance Management and Field Service Management offers tools for better financial planning and customer service. Gary Wiessinger, SVP of Application Development at Oracle NetSuite, emphasized that these innovations will help businesses adapt swiftly to evolving market demands.

SAP introduces AI-powered visual inspections to enhance asset management

SAP has unveiled AI-enabled visual inspection capabilities within its Asset Performance Management (APM) platform, aiming to streamline maintenance processes for industries reliant on heavy equipment, such as manufacturing, utilities, and transportation. This innovation allows organizations to integrate camera and drone imagery with analytical tools, partially automating inspection tasks and reducing the need for manual assessments. SAP's approach emphasizes integration with existing AI models through its partner ecosystem, enabling companies to leverage current investments. Additionally, the system incorporates feedback mechanisms where technicians can validate AI analyses, fostering continuous improvement and increased trust in automated processes.

AkiraBot exploits AI to spam Over 420,000 websites, bypassing CAPTCHA protections

Cybersecurity researchers have uncovered AkiraBot, an AI-powered spam tool that has targeted more than 420,000 websites, successfully spamming at least 80,000 since September 2024. Utilizing OpenAI's GPT-4o-Mini model, AkiraBot crafts customized outreach messages to promote dubious SEO services like Akira and ServicewrapGO. The bot employs proxy services to mimic legitimate users, effectively bypassing CAPTCHA systems such as hCAPTCHA, reCAPTCHA, and Cloudflare Turnstile. It maintains logs of its activities and shares success metrics via a Telegram channel. In response to these findings, OpenAI has disabled the API key and associated assets used by the threat actors.

Cybercriminals manipulate search results to direct users to malicious websites

Cybercriminals are increasingly manipulating search engine results to lead unsuspecting users to malicious websites, exploiting the trust users place in top search results. This tactic, known as SEO poisoning, involves manipulating search engine algorithms to artificially boost a site’s ranking, while malicious search advertisements place dangerous content at the top of search results through paid advertising platforms. Once users click on these compromised results, they are often directed to counterfeit websites that closely mimic legitimate services, designed to harvest sensitive information or distribute malware. Despite efforts by companies like Google to block malicious ads, these deceptive tactics continue to pose significant threats to online security.