SpyCloud, a leader in identity threat protection, has published its 2026 Forecast Report, offering insights into cybersecurity threats expected in 2027. The report draws from expertise across SpyCloud’s security research teams and a survey of 50 professionals from large enterprises. It highlights stolen sessions, API keys, and trusted vendor connections as potential drivers of significant cyberattacks in the coming year.
AI’s Dual Role in Cybersecurity
Artificial Intelligence (AI) plays a dual role in cybersecurity, accelerating both attacks and defenses. Organizations are rapidly adopting AI, but governance lags behind, giving attackers an edge to automate phishing and enhance malware. SpyCloud’s Aurora Johnson notes that attackers increasingly exploit trusted access points such as employee sessions and vendor links, emphasizing the need for more robust security measures beyond multifactor authentication.
Emerging Threats in 2027
SpyCloud experts predict that attackers will bypass multifactor authentication by compromising devices or sessions. Phishing tactics will evolve, incorporating methods like device code phishing and adversary-in-the-middle attacks. Furthermore, attackers will increasingly target API keys and tokens associated with AI tools, posing a threat to enterprise systems. SpyCloud has already detected over 127,000 instances of one particular infostealer, highlighting the growing value of non-human identity credentials.
The report also forecasts that trusted access points, such as vendor integrations and insider threats, will be major avenues for damaging breaches. Connections authorized long ago could still hold administrative privileges, making them prime targets. With 68% of surveyed security professionals reporting breaches via vendors or contractors, the risk is substantial.
Fraud and Identity Challenges
As AI-generated identities become more sophisticated, fraud detection will become more challenging. Cybercriminals will continue to create convincing fake identities using stolen data and AI-generated visuals. Although only a small percentage of survey respondents have reported such incidents, SpyCloud anticipates a rise in these cases as AI agents are increasingly used to automate account creation.
Additionally, AI agents themselves will become targets of phishing attacks. These agents possess similar access levels to human employees, making them vulnerable. Early indicators include malicious activities on AI marketplaces designed to spread malware.
Outlook for 2027 and Beyond
SpyCloud’s report suggests that while fully autonomous attacks may capture headlines in 2027, traditional attacks enhanced by AI will inflict the most damage. The focus will shift towards stealing tokens, keys, and service accounts that connect various systems. Joe Roosen from SpyCloud stresses the growing importance of protecting these non-human identities as SaaS platforms proliferate.
SpyCloud’s Jason Lancaster underscores the necessity for increased visibility and remediation capabilities to prevent exploits of compromised access. By leveraging over a trillion recaptured identity assets, SpyCloud aids organizations in mitigating risks associated with session hijacking and account takeovers. To delve deeper into SpyCloud’s findings and enhance your organization’s cybersecurity posture, visit spycloud.com.
SpyCloud remains a pivotal player in transforming recaptured darknet data to counter cybercrime, safeguarding identities across diverse sectors. As the digital landscape evolves, maintaining vigilance against emerging threats remains crucial for businesses and consumers alike.
