#artificialIntelligenceSecurity
Over 18 Million Files Leaked in AI Aggregator Breach, Raising Privacy Concerns #AIDataExposure #AISecurity #ArtificialIntelligenceSecurity
Over 18 Million Files Leaked in AI Aggregator Breach, Raising Privacy Concerns
There has been a major data exposure at the AI aggregator 302.ai that has resulted in millions of documents containing private user conversations, personal information and confidential business material becoming publicly accessible.  Cybernews cybersecurity researchers discovered an unsecured Microsoft Azure storage bucket belonging to this platform, which exposed approximately 22.72 terabytes of information. There were 17.6 million files in the exposed archive, including images, videos, audio recordings, documents and chat histories. The storage bucket was publicly accessible without password protection or encryption, resulting in anyone able to discover its location having access to the information stored there.  302.ai is a platform that facilitates the interaction between users and a variety of artificial intelligence services, enabling customers to access multiple large language models on one platform. In addition to ChatGPT, Claude, Grok, DeepSeek, and Gemini, the company supports ChatGPT, Claude, Grok, and Gemini. This arrangement permits users to switch between artificial intelligence models without having to access each provider separately.  According to Cybernews, the exposed storage contained approximately 144,000 JSON-formatted conversation logs. Users' prompts, AI-generated responses, models used to process requests, timestamps, token consumption costs, and other related metadata were part of these records. As part of the conversations, users entered personally identifiable information as they interacted with the platform.  The records included details about their birthdates, genders, names, contact information, and addresses, among other personal information. A combination of such information and conversation histories and uploaded documents can reveal more details about individuals than their ordinary chatbot activity. The exposure was more extensive than text-based communications. The archive contained more than 9 terabytes of images, 7.7 terabytes of videos, and another 5.6 terabytes of audio files. The researchers also discovered over 26,000 ZIP archives containing conversation summaries, inputs from users, AI responses, and attachments to individual sessions. According to reports, a number of documents within the storage were related to large language model development research projects and materials reportedly submitted for private analysis. It is evident from the findings that the information exposed was not confined to individuals, but included information related to business activities as well.  An especially sensitive discovery involved multiple conversations with a chatbot that provided customer support for a prominent manufacturer of sex dolls. These records illustrate the increasing use of third-party artificial intelligence platforms to manage enterprise customer interactions, resulting in potentially exposing confidential and personal information to external storage environments. Cybernews failed to identify the manufacturer or determine whether all of its customer-service records were exposed.  Despite the discovery, other concerns have been raised regarding the handling of sensitive customer conversations by third-party artificial intelligence services. As opposed to a confirmed cyberattack, researchers stated that the exposure was likely the result of an accidental storage configuration error. Nonetheless, the investigation did not determine whether any unauthorised parties accessed the files prior to the security measures being implemented or if the technical cause could be determined. The exposed Azure storage was discovered on August 2, 2026, as reported by Cybernews. A team of researchers contacted 302.ai on August 11 and observed that public access had been disabled by September 8. Researchers found no evidence of malicious actors downloading the information, however it was not impossible for unknown parties to obtain access. Exposed AI Conversations Could Enable Targeted Cyberattacks Cybercriminals might be able to conduct targeted phishing and social engineering campaigns if these information is obtained by malicious individuals. As a result of chat histories revealing personal details, business relationships, ongoing projects, customer concerns, and other information, attackers are able to make fraudulent communications appear more convincing by using this information.  Conversations between employees, customers or senior management could be particularly useful for impersonation attempts. An attacker could also use information contained in uploaded documents and interactions with chatbots to identify potential targets and gain an understanding of internal company operations. The incident demonstrates the security risks associated with third-party AI aggregators.  In addition to providing access to models developed by major technology companies, platforms such as 302.ai also maintain their own infrastructure to handle prompts, conversation histories, and uploaded files. Generally, the security protections offered by the underlying AI providers do not extend to third-party intermediaries.  Businesses Urged to Review AI Data Handling Practices For organizations that rely on artificial intelligence platforms to process customer inquiries, internal documents, and other confidential information, the exposure raises significant concerns. Companies using 302.ai should identify which employees and departments have used the service and whether sensitive company information or customer information has been shared.  A security team should also review all available conversation histories and uploaded files for potential vulnerabilities in credentials, API keys, and confidential documents. Credentials discovered in submitted materials should be rotated, while suspicious account activity should be investigated. In light of the incident, it is important to evaluate third-party AI vendors before incorporating them into business workflows. Providers should be examined to determine how they restrict access to storage, protect uploaded files, separate customer information, and manage data retention and deletion. 302.ai is developed by Univerads Technology Limited, a Hong Kong-based company founded in 2020. This exposure illustrates how weaknesses in intermediary platforms can result in sensitive information being compromised, even when the AI models accessed through those services are operated by established technology companies. It is clear from the 302.ai exposure that there are growing security concerns associated with third-party artificial intelligence platforms and sensitive data they handle. Businesses are increasingly integrating artificial intelligence tools into their operations; however, preventing confidential information from being made public requires strong data protection measures, regular security assessments, and thorough vendor evaluations.
dlvr.it
November 5, 2026 at 3:29 PM
October 14, 2026 at 10:59 AM
Cybersecurity Falls Behind as Threat Scale Outpaces Capabilities #AIDrivenThreats #ArtificialIntelligenceSecurity #CyberGovernance
Cybersecurity Falls Behind as Threat Scale Outpaces Capabilities
Cyber defence is entering its 2026 year with the balance of advantage increasingly being determined by speed rather than sophistication. With the window between intrusion and impact now measured in minutes rather than days instead of days, the advantage is increasingly being gained by speed.  As breakout times fall below an hour and identity-based compromise replaces malware as the dominant method of entry into enterprise environments, threat actors are now operating faster, quieter, and with greater precision than ever before.  By making use of artificial intelligence, phishing, fraud, and reconnaissance can be executed at unprecedented scales, with minimal technical knowledge, which is a decisive accelerator for the phishing, fraud, and reconnaissance industries. As a result of the commoditization, automation, and availability of capabilities once requiring specialized skills, they have lowered the barrier to entry for attackers dramatically.  There is an increased threat of "adaptive, fast-evolving threats" that organizations must deal with, and one of the main factors that has contributed to this is the rapid and widespread adoption of artificial intelligence across both offensive and defensive cyber operations. Moody's Ratings describes this as leading to a "new era of adaptive, fast-evolving threats".  A key reality for chief information security officers, boards of directors, and enterprise risk leaders is highlighted in the firm's 2026 Cyber Risk Outlook: Artificial intelligence isn't just another tool in cybersecurity, but is reshaping the velocity, scale, and unpredictability of cyber risk, impacting both the management, assessment, and governance of cyber risks across a broad range of sectors.  While years have been spent investing and innovating in enterprise security, the failure of enterprise security rarely occurs as a consequence of a lack of tools or advanced technology; rather, failure is more frequently a result of operating models that place excessive and misaligned expectations on human defenders, forcing them to perform repetitive, high-stakes tasks with fragmented and incomplete information in order to accomplish their objectives.  Modern threat landscapes have changed considerably from what was originally designed to protect static environments to the dynamic environment the models were built to protect. Attack surfaces are constantly changing as endpoints change their states, cloud resources are continually being created and retired, and mobile and operational technologies are continuously extending exposures well beyond traditional perimeters.  There has been a gradual increase in threat actors exploiting this fluidity, putting together minor vulnerabilities one after another, confident that eventually defenders will not be able to keep up with them.  A huge gap exists between the speed of the environment and the limits of human-centered workflows, as security teams continue to heavily rely on manual processes for assessing alerts, establishing context, and determining when actions should be taken.  Often, attempts to remedy this imbalance through the addition of additional security products have compounded the issue, increasing operational friction, as tools overlap, alert fatigue is created, and complex handoffs are required.  Despite the fact that automation has eased some of this burden, it still has to do with human-defined rules, approvals, and thresholds, leaving many companies with security programs that may appear sophisticated at first glance but remain too slow to respond rapidly, decisively, in crisis situations. Various security assessments from global bodies have reinforced the fact that artificial intelligence is rapidly changing both cyber risk and its scale. In a report from Cloud Security Alliance (CSA), AI has been identified as one of the most important trends for years now, with further improvements and increased adoption expected to accelerate its impact across the threat landscape as a whole. It is cautioned by the CSA that, while these developments offer operational benefits, malicious actors may also be able to take advantage of them, especially through the increase of social engineering and fraud effectiveness.  AI models are being trained on increasingly large data sets, making their output more convincing and operationally useful, and thus making it possible for threat actors to replicate research findings and translate them directly into attack campaigns based on their findings. CSA believes that generative AI is already lowering the barriers to more advanced forms of cybercrime, including automated hacking as well as the potential emergence of artificial intelligence-enabled worms, according to the organization.  It has been argued by David Koh, Chief Executive of the Cybersecurity Commissioner, that the use of generative artificial intelligence brings to the table a whole new aspect of cyber threats, arguing that attackers will be able to match the increased sophistication and accessibility with their own capabilities.  Having said that, the World Economic Forum's Global Cybersecurity Outlook 2026 is aligned closely with this assessment, whose goal is to redefine cybersecurity as a structural condition of the global digital economy, rather than treating it as a technical or business risk. According to the report, cyber risk is the result of convergence of forces, including artificial intelligence, geopolitical tensions, and the rapid rise of cyber-enabled financial crime.  A study conducted by the Dublin Institute for Security Studies suggests that one of the greatest challenges facing organizations is not the emergence of new threats but rather the growing inadequacy of existing business models related to security and governance.  Despite the WEF's assessment that the most consequential factor shaping cyber risk is the rise of artificial intelligence, more than 94 percent of senior leaders believe that they can adequately manage the risks associated with AI across their organizations. However, fewer than half indicate that they feel confident in their ability to manage these risks. According to industry analysts, including fraud and identity specialists, this gap underscores a larger concern that artificial intelligence is making scams more authentic and scaleable through automation and mass targeting. These trends, taken together, indicate that organizations are experiencing a widening gap between the speed at which cyber threats are evolving and their ability to identify, respond, and govern them effectively as a result.  Tanium offers one example of how the transition from tool-centered security to outcome-driven models is taking shape in practice, reflecting a broader shift from tool-centric security back to outcomes-driven security. This change in approach exemplifies a growing trend of security vendors seeking to translate these principles into operational reality.  In addition to proposing autonomy as a wholesale replacement for established processes, the company has also emphasized the use of real-time endpoint intelligence and agentic AI as a method of guiding and supporting decision-making within existing operational workflows in order to inform and support decision-making.  The objective is not to promote a fully autonomous system, but rather to provide organizations with the option of deciding at what pace they are ready to adopt automation. Despite Tanium leadership's assertion that autonomous IT is an incremental journey, one involving deliberate choices regarding human involvement, governance, and control, it remains an incremental journey.  The majority of companies begin by allowing systems to recommend actions that are manually reviewed and approved, before gradually permitting automated execution within clearly defined parameters as they build confidence in their systems.  Generally, this measured approach represents a wider understanding of the industry that autonomous systems scale best when they are integrated directly into familiar platforms, like service management and incident response systems, rather than being added separately as a layer.  Vendors are hoping that by integrating live endpoint intelligence into tools like ServiceNow, security teams can shorten response times without requiring them to reorganize their operations. In essence, this change is a recognition that enterprise security is about more than eliminating complexity; it's about managing it without exhausting the people who need to guard increasingly dynamic environments.  In order to achieve effective autonomy, humans need not be removed from the loop, but rather effort needs to be redistributed. It has been observed that computers are better suited for continuous monitoring, correlation, and execution at scale, while humans are better suited for judgment, strategic decision-making, and exceptional cases, when humans are necessary.  There is some concern that this transition will not be defined by a single technological breakthrough but rather by the gradual building up of trust in automated decisions. It is essential for security leaders to recognize that success lies in creating resilient systems that are able to keep up with the ever-evolving threat landscape and not pursuing the latest innovation for its own sake.  Taking a closer look ahead, organizations are going to realize that their future depends less on acquiring the next breakthrough technology, but rather on reshaping how cyber risk is managed and absorbed by the organization. In order for security strategies to be effective in a real-world environment where speed, adaptability, and resilience are as important as detection, they must evolve. Cybersecurity should be elevated from an operational concern to a board-level discipline, risk ownership should be aligned to business decision-making, and architectures that prioritize real-time visibility and automated processes must be prioritized.  Furthermore, organizations will need to put more emphasis on workforce sustainability, and make sure that human talent is put to the best use where it can be applied rather than being consumed by routine triage.  As autonomy expands, both vendors and enterprises will need to demonstrate that they have the technical capability they require, as well as that they are transparent, accountable, and in control of their business.  Despite the fact that AI has shaped the environment, geopolitics has shaped economic crime, and economic crime is on the rise, the strongest security programs will be those that combine technological leverage with disciplinary governance and earned trust.  It is no longer simply necessary to stop attacks, but rather to build systems and teams capable of responding decisively in a manner that is consistent with the evolving threat landscape of today.
dlvr.it
January 17, 2026 at 5:18 PM
UK strengthens AI defenses to resist Russian cyber threats. Never doubt, the UK stands resilient! #ArtificialIntelligenceSecurity

https://www.theregister.com/2024/11/26/uk_ai_security/
Britain boffins gear up for AI warfare with Russia
Be in no doubt: the UK and others in this room are watching Russia
www.theregister.com
November 26, 2024 at 11:14 AM