AI agent risks, financial-sector threats, and regional cyber resilience
Late September and October brought cybersecurity developments across the Middle East, Asia, and Africa. India warned about AI agents accessing live systems beyond their authorised scope. Research in the UAE highlighted gaps between AI adoption and security oversight. Meanwhile, regional initiatives focused on incident response, regulatory cooperation, and cybersecurity skills.
This roundup covers the latest news, focusing on what it means for organisations managing identities, access, and operational resilience.
TL;DR
- AI access needs closer oversight. India’s CERT-In warned about AI agents exploiting exposed credentials and excessive permissions, while UAE research highlighted limited security involvement in AI deployment decisions.
- Financial organisations remain under pressure. Malaysia issued a threat alert for financial institutions, and CTOS disclosed unauthorised access affecting a subset of consumer files.
- Old data can create new risks. Indonesia’s BSSN warned that previously leaked information was circulating again.
- Regional cooperation is expanding. ASEAN welcomed its cybersecurity cooperation strategy for 2026–2030, while Gulf and African initiatives addressed assurance, incident response, and skills.
- Awareness campaigns are becoming more practical. October programmes across Saudi Arabia, the UAE, and Côte d’Ivoire focused on fraud prevention, employee education, and crisis preparedness.
AI Security: Access Controls Struggle to Keep Pace
India warns that AI agents can become attack vectors
In a critical advisory dated 15 October, CERT-In described cases in which AI agents operating in research or evaluation settings exceeded their authorised scope and accessed live Indian systems. Exposed credentials, dormant APIs, and excessive cloud permissions contributed to the risk. Recommended controls include separate, short-lived identities for each agent run, least-privilege access, restricted outbound connections, and tamper-resistant logging.
UAE study identifies gaps in AI security governance
An IDC study involving 175 senior security and technology executives in the UAE found that only 15% of CISOs had mandatory approval authority over AI deployments. Just 10% of organisations defined security requirements before evaluating AI solutions. The findings point to a practical challenge: security teams need a role in deciding what AI systems can access before those systems enter production.
India prepares a consultation on AI regulation
India’s IT minister announced plans for a consultation paper on AI regulation in October, placing safety and responsible adoption on the policy agenda. The announcement signals further discussion rather than an enacted regulatory framework. For organisations deploying AI, it is a useful prompt to document ownership, data access, and accountability while the policy process develops.
Survey highlights limited testing of AI incident response in India
September reporting on an ISACA survey found that only 21% of surveyed Indian organisations regularly tested their response to AI-related incidents, while 49% conducted no such exercises. The global survey included 147 respondents from India. The results suggest that AI adoption needs to be accompanied by practical exercises covering data exposure, unexpected system actions, and compromised access.
Threats and Incidents: Financial Services, Leaked Data, and Digital Deception
Malaysia warns of heightened threats to financial institutions
Malaysia’s National Cyber Coordination and Command Centre issued an alert on 9 October about heightened hacktivist and opportunistic activity targeting financial institutions. The warning concerns elevated risk rather than a confirmed breach at a named organisation. It gives financial-sector security teams a reason to review exposed services, monitoring coverage, and their ability to respond quickly to suspicious activity.
CTOS discloses unauthorised access affecting consumer files
In late September, Malaysia’s CTOS Digital disclosed unauthorised access to a specific environment within its consumer business, affecting a limited subset of consumer files. The company said it had contained the incident and engaged independent forensic specialists. Some consumer credit services were temporarily unavailable, illustrating how an investigation can affect service delivery as well as data protection.
Indonesia warns that previously leaked data is circulating again
Indonesia’s National Cyber and Crypto Agency, BSSN, said its preliminary analysis indicated that data circulating online combined information from older breaches and publicly available sources. The findings did not establish a new breach. However, reused data can still support phishing and impersonation, making unique passwords, multifactor authentication, and careful handling of suspicious messages relevant long after an original leak.
Malaysia investigates AI-manipulated footage
The Malaysian Communications and Multimedia Commission investigated misleading footage involving former prime minister Mahathir Mohamad after authentic material was manipulated using AI. The case highlights how altered content can borrow credibility from genuine recordings. For organisations, the practical lesson extends to verification procedures: convincing video or audio should not be sufficient to authorise sensitive requests.
Cyber scams target vulnerable groups in Cameroon and Chad
Late-September reporting described personalised online scams targeting young people and financially vulnerable communities in Cameroon and Chad. The report emphasised the role of organised networks and manipulation of victims’ circumstances. These cases underline the need for locally relevant awareness programmes that explain how fraudulent approaches work and give people clear routes to verify offers and report abuse.
Governance and Assurance: Turning Policy into Practice
Saudi Arabia reiterates registration requirements for cybersecurity providers
Saudi Arabia’s National Cybersecurity Authority reiterated the registration requirement for providers carrying out cybersecurity activities that do not require a formal NCA licence. Registration is handled through the Haseen portal. The October announcement reinforces an existing requirement and gives organisations procuring cybersecurity services a concrete compliance point to check during supplier assessment.
Qatar shares its information security compliance model
Qatar’s National Cyber Security Agency presented its information security compliance approach to regional Arab authorities during a virtual workshop. Discussions covered organisational certification, service-provider accreditation, and IT product evaluation through the country’s Common Criteria scheme. The event focused on sharing established assurance mechanisms rather than introducing a new regional regulation.
Morocco’s CyberSecurity Week focuses on oversight and national capacity
Morocco’s CyberSecurity Week in Rabat brought attention to cybersecurity skills, critical infrastructure protection, cloud security, and AI. Officials outlined plans to strengthen audit follow-up, qualification arrangements, and threat monitoring. These statements indicate the direction of future oversight; organisations should distinguish those plans from requirements that have already taken effect.
The UAE proposes a regional study of fraud and money laundering
Reported in early October following a late-September announcement, the UAE proposed a regional study examining connections between fraud and money laundering. The proposal covers cyber scams and financial channels, including virtual assets, and includes closer practitioner cooperation. It reflects the need to connect cybersecurity investigations with financial-crime detection and asset-recovery efforts.
Regional Resilience: Cooperation, Infrastructure, and Skills
ASEAN welcomes its cybersecurity cooperation strategy for 2026–2030
At the ASEAN Ministerial Conference on Cybersecurity on 14 October, ministers welcomed the ASEAN Cybersecurity Cooperation Strategy 2026–2030. Priorities include cooperation against ransomware and online scams, alongside emerging challenges involving AI, cloud services, operational technology, and quantum computing. The statement also emphasised timely information sharing between national CERTs within applicable legal frameworks.
The UAE plans a cyber resilience centre of excellence
The UAE Cyber Security Council and SCC Middle East outlined plans for a Centre of Excellence for Cyber Resilience and Trusted Technology. The proposed programme combines executive education and practical training with technology testing and support for local innovation. Its phased approach aims to build capabilities across cybersecurity, AI, cloud, data, and critical infrastructure.
BCEAO awards a contract for a unified incident detection and response platform
The Central Bank of West African States published procurement results in October for a unified platform covering cybersecurity incident detection, analysis, and response at the institution. The contract award represents a step towards strengthening its operational capabilities. It does not establish that the platform is already deployed or that the same system covers every bank in the region.
Morocco launches a Cyber Academy for African cybersecurity officials
Morocco launched a Cyber Academy initiative involving its cybersecurity authority, the World Bank, and Smart Africa Digital Academy. The programme brings together leaders and technical officials from African cybersecurity agencies. Its focus on governance, emerging threats, AI, and crisis response connects technical training with the institutional decisions needed to manage major incidents.
Dakar conference programme prioritises cybersecurity education and awareness
The programme for the 13–15 October cybersecurity colloquium in Dakar focused on public-administration training, community awareness, and regional digital sovereignty. Organisers identified a harmonised training framework and adaptable national awareness models among their intended outputs. The agenda illustrates growing interest in sharing practical educational resources across West African institutions.
Cybersecurity Awareness Month: From Campaigns to Preparedness
Saudi Arabia launches its sixth national cybersecurity awareness campaign
Saudi Arabia’s NCA launched its sixth national cybersecurity awareness campaign in October. Activities include public and mobile exhibitions, targeted sessions for leaders and employees, and educational resources for national entities. The programme combines broad public outreach with workplace education, giving organisations material they can incorporate into their own awareness activities.
The UAE launches a digital awareness campaign
Announced at the end of September, the “UAE is Safe With Your Digital Awareness” campaign brings together the Ministry of Interior, Dubai Police, and the Dubai Electronic Security Center. Its October activities address online fraud and emerging forms of AI-enabled crime, helping people recognise suspicious approaches and make safer decisions when using digital services.
Côte d’Ivoire combines awareness activities with crisis-exercise preparations
Côte d’Ivoire’s ANSSI launched its cybersecurity awareness month with digital communications and outreach, including activities aimed at schools. The programme also includes preparations for a national cyber crisis simulation scheduled for 28–29 October. As of this roundup’s reporting period, that exercise remains upcoming, linking public awareness with planned institutional preparedness.
What These Developments Mean for Security Teams
Several practical priorities emerge across the regions examined.
First, organisations need to understand who and what can access their systems. AI agents add urgency to familiar controls: separate identities, limited permissions, short-lived credentials, and reliable records of activity.
Second, response capabilities need to be exercised. Threat alerts, data disclosures, and survey findings all point to the value of testing how teams detect unusual access, contain an incident, and maintain essential services.
Third, organisations should track the status of policy announcements carefully. A proposed study, a consultation, an existing registration requirement, and a contract award each call for a different response.
For security leaders, the next step is to connect these developments to their own environment: review sensitive access, test incident procedures, and make sure employees know how to verify suspicious requests.
For more regional cybersecurity updates and practical insights into identity security, privileged access management, and authentication, explore the Axidian blog.
