An agreement signed between Iran and Russia last week outlines commitments to enhance military, security, cyber and technological cooperation between the two nations. The comprehensive strategic partnership agreement, signed in Moscow by Russian President Vladimir Putin and Iranian President Masoud Pezeshkian, seeks to deepen bilateral relations and includes specific provisions for cooperation in cybersecurity and internet regulation.
The agreement aims to counter the use of information and communication technologies for criminal activities and includes plans to exchange expertise on managing national internet infrastructure. The text also adds that two countries will ‘promote the establishment of a United Nations-led system for ensuring international information security and the creation of a legally binding regime for the prevention and peaceful resolution of conflicts, based on the principles of sovereign equality and non-interference in the internal affairs of states’.
The agreement emphasises strengthening sovereignty and state-centric approach to international information security and internet governance. Other key commitments on cybersecurity also include:
Expanding joint efforts to combat the criminal misuse of ICTs, exchanging expertise, and promoting sovereignty in the international information domain.
Advocating for the internationalization of internet governance, equal rights for states in managing internet segments, and rejecting limitations on national sovereignty in regulating and securing the internet.
Enhancing sovereignty through regulating global ICT companies, sharing expertise on internet management, developing ICT infrastructure, and advancing digital development.
The Cybersecurity and Infrastructure Security Agency (CISA) and the Federal Bureau of Investigation (FBI) have released detailed guidance aimed at software manufacturers to enhance security across the product lifecycle. This document applies to all software products and services, including on-premises software, cloud services, Software as a Service (SaaS), operational technology (OT), and embedded systems. While non-binding, the guidance encourages manufacturers to adopt secure-by-design principles and reduce risks for their customers by avoiding specific bad practices.
The guidance reflects feedback from 78 public comments and introduces three new bad practices:
Using known insecure or outdated cryptographic functions.
Hardcoded credentials.
Insufficient product support periods.
Updates also include:
Enhanced context on memory safety and multi-factor authentication (MFA), particularly for OT products.
New examples of actions to prevent SQL injection and command injection vulnerabilities.
Clear timelines for addressing Known Exploited Vulnerabilities (KEVs).
Some of the recommendation actions to software manufacturers specifically address the critical infrastructure protection. For instance, Software manufacturers are urged to:
Prevent command injection vulnerabilities: Use library functions, sanitize inputs with restrictive allowlists, and delineate command inputs.
Eliminate default passwords: Implement instance-unique, random passwords; enforce secure credentials during setup; and support phishing-resistant MFA.
Patch Known Exploited Vulnerabilities (KEVs): Issue free patches within 30 days of a KEV’s inclusion in CISA’s catalog and communicate risks to users.
Support Open Source Software(OSS): Contribute responsibly and sustainably to open-source projects relied upon.
By following this guidance, manufacturers signal their commitment to customer security and contribute to a safer software ecosystem.
OpenAI has launched an advanced AI agent, Operator, designed to automate repetitive online tasks such as ordering groceries or booking restaurants. Operator uses its own browser to interact with buttons, menus, and text fields, removing the need for custom API integrations.
The tool operates independently once enabled, but it is designed to seek user approval for actions involving sensitive information like logins or payments. Its functionality relies on OpenAI’s GPT-4o vision model and advanced reasoning capabilities, allowing it to self-correct mistakes or transfer control back to users when needed.
Operator is currently in a research preview phase, available exclusively to US-based ChatGPT Pro subscribers. Feedback from early adopters will shape its future development before it becomes accessible to more users.
OpenAI is collaborating with companies including DoorDash, Uber, and StubHub to align Operator with real-world needs and industry norms, ensuring practicality and reliability for diverse applications.
US President Donald Trump has signed an executive order aimed at solidifying the country’s dominance in artificial intelligence. The directive includes creating an Artificial Intelligence Action Plan within 180 days to promote economic competitiveness, national security, and human well-being. The White House confirmed this initiative as part of efforts to position the nation as a global AI leader.
Trump has also instructed his AI and national security advisers to dismantle policies implemented by former President Joe Biden. Among these is a 2023 order requiring AI developers to submit safety test results to the government for systems with potential risks to national security, public safety, or the economy.
Biden’s policies aimed to regulate AI development under the Defence Production Act to minimise risks posed by advanced technologies. Critics argue the approach imposed unnecessary constraints, while supporters viewed it as a safeguard against potential misuse of AI.
The latest move reflects Trump’s broader strategy to reshape the nation’s AI framework, focusing on economic growth and innovation while rolling back measures seen as restrictive.
Rivian, the US electric vehicle maker, and Volkswagen are in talks with other automakers about supplying them with software and electrical architecture through their joint venture. This collaboration, which began in November with Volkswagen’s $5.8 billion investment, aims to integrate advanced electrical infrastructure and Rivian’s software technology into both companies’ future EVs. Rivian’s streamlined vehicle architecture, which reduces weight and manufacturing complexity, also allows for over-the-air software updates, an area where traditional automakers have struggled to catch up.
Rivian‘s Chief Software Officer, Wassym Bensaid, revealed that other automakers are interested in the joint venture’s technology, though he declined to name them or provide details on the ongoing discussions. The venture is a key opportunity for established automakers to quickly access the technology they have long sought to develop themselves. For Rivian, the partnership provides higher volumes, better supplier deals, and a chance to reduce costs, especially important as EV demand slows.
Rivian focuses on launching its smaller, more affordable R2 SUV by 2027, while also expanding the integration of its technology into Volkswagen’s other brands. With increasing interest from additional OEMs, the joint venture is poised to become a significant player in the global EV market, particularly in the West, alongside Tesla. Analysts suggest the partnership helps Rivian address its capital concerns and positions it as a key player in the transition to software-defined vehicles.
Stargate, a new joint venture formed by OpenAI, SoftBank, and Oracle, aims to build data centres across the US to support the growing demands of AI. According to a report by the Financial Times on Thursday, these data centres will be dedicated solely to OpenAI, the company behind the popular ChatGPT. The collaboration between these tech giants underscores the increasing importance of robust infrastructure to power the next wave of AI innovation.
The exclusive focus on OpenAI’s needs comes when AI technologies rapidly expand, with the demand for high-performance computing capabilities soaring. The partnership will allow OpenAI to scale its operations and provide the necessary computing power for its cutting-edge AI models. As companies worldwide race to develop more advanced AI tools, the infrastructure provided by Stargate is expected to play a crucial role in supporting the next generation of AI services.
Oracle and SoftBank’s involvement brings significant expertise in cloud infrastructure and global telecom, making the venture a powerful alliance in the competitive AI landscape. The project highlights the growing intersection of cloud computing, data storage, and AI as companies like OpenAI push the boundaries of what AI can achieve.
The founder and former CEO of GameOn, an AI startup in San Francisco, has been indicted for orchestrating a six-year-long fraud scheme that allegedly defrauded investors and the company out of over $60 million. Alexander Beckman, 41, faces 23 criminal charges, while his wife, Valerie Lau Beckman, 38, who worked as a lawyer for the company, is charged with 16 counts, including obstruction. Both have pleaded not guilty. The US Securities and Exchange Commission has also filed civil charges against the couple.
Beckman is accused of deceiving investors by inflating the company’s financial status, including fabricating fake customer relationships, overstating revenue, and creating fraudulent bank statements and audit reports. He allegedly went as far as impersonating individuals to share false information. Meanwhile, Lau Beckman allegedly assisted her husband by providing authentic audit reports to help fabricate false documents and delete critical files after an investigation began.
The Beckmans are also accused of misusing investor funds for personal expenses, including purchasing a luxury home, vehicles, and covering costs for their wedding. The fraudulent activities reportedly continued up until Beckman’s resignation as CEO in July 2024. GameOn, which has since been rebranded as On Platform, eventually admitted to the financial discrepancies and laid off most of its employees.
The case underscores the need for integrity in the tech industry, particularly within startups, as federal prosecutors emphasise that fraud cannot fuel innovation.
South Sudan has suspended access to social media platforms for at least 30 days following violent riots triggered by videos allegedly showing the killings of South Sudanese nationals in Sudan’s El Gezira state. The decision, announced by the National Communications Authority on Wednesday, aims to curb the spread of extreme content and prevent further unrest. Mobile operators MTN South Sudan and Zain confirmed that platforms like Facebook and TikTok would be inaccessible for up to 90 days.
The riots, which erupted in the capital, Juba and other cities, led to the deaths of at least 16 Sudanese nationals. Angry youths looted shops, vandalised property, and burned homes belonging to Sudanese nationals, believing Sudan’s military and its allies were involved in the El Gezira killings. South Sudanese authorities have condemned the violence, urging calm and restraint.
The Sudanese army has also criticised what it described as ‘individual violations’ in El Gezira. The social media ban is part of a broader effort to restore order and prevent further acts of retaliation, as tensions remain high between the neighbouring nations.
The US government is introducing the Cyber Trust Mark, a new security certification aimed at safeguarding smart home devices against cyber threats. Launching later this year, the programme will provide consumers with a clear indicator of which gadgets meet strict cybersecurity standards set by the National Institute of Standards and Technology (NIST). Devices such as smart cameras, fitness trackers, and baby monitors are among those eligible for the label.
To qualify, manufacturers must implement measures such as strong default passwords, software updates, and data protection protocols. Shoppers can also scan a QR code accompanying the label for detailed security information, including tips on setup and maintenance. The initiative comes in response to the rising threat of hackers targeting home networks, with the average US household now owning over 20 connected devices.
Retail giants like Amazon and Best Buy are backing the programme, highlighting compliant products to help consumers make informed choices. While the Cyber Trust Mark focuses on wireless Internet of Things (IoT) gadgets, some devices, including medical equipment, cars, and wired products, are excluded. The scheme marks a significant step toward protecting homes from cybercrime as digital threats continue to grow.
What impact this label will have on consumer habits remains to be seen, but it’s already drawing support from major tech firms like Google and Samsung, signalling a collective move towards better digital security.
Meta users in the US are experiencing an unusual phenomenon where they are being automatically re-followed by the accounts of President Donald Trump, Vice President JD Vance, and first lady Melania Trump. The issue emerged after users intentionally unfollowed these accounts following the administration’s transition. Feedback from users, including actress Demi Lovato and comedian Sarah Colonna, highlighted frustration over the inability to maintain their choice to unfollow prominent political figures.
Upon the change of administration, official White House social media accounts are supposed to transition smoothly to the new leaders. While Meta’s communications director Andy Stone acknowledged that followers from the Biden administration were carried over to Trump’s accounts, he confirmed that users were not being forced to re-follow these profiles. Stone suggested that delays in processing follow and unfollow requests might contribute to the confusion experienced by users.
Many individuals reported recurrent issues despite efforts to unfollow the accounts multiple times, raising questions about the underlying technicalities involved. Users are expressing concerns over privacy and choice in the use of social media platforms, as the ability to curate their feeds appears compromised. However, this automatic re-following could reflect broader implications for user control in digital spaces.
As Meta has yet to release a detailed response to the reported glitch, users continue to voice their concerns across multiple platforms. The situation underscores an ongoing need for clarity and assurance regarding user preferences in social media interactions, especially during a politically sensitive time.
