Artificial Intelligence
OpenAI Announces GPT-5.4-Cyber To Match Rivals’ AI model
OpenAI has unveiled GPT-5.4-Cyber, a specialized AI built for defensive cybersecurity, shortly after Anthropic introduced its Mythos model. The tool is being released through a tiered “Trusted Access” program, giving verified experts the ability to detect and fix software vulnerabilities more effectively. As cyber threats continue to grow globally, this marks an important step in using AI to safeguard critical systems. For Australia, these advancements provide new opportunities for organizations to strengthen their defenses with cutting-edge, security-focused AI tools.
OpenAI has introduced a new tool called GPT-5.4-Cyber. It’s a specialised version of their AI, built specifically for cybersecurity experts to help protect websites and applications from digital threats.
Key Highlights
- OpenAI has introduced GPT-5.4-Cyber for security-focused work
- It was released just one week after rival Anthropic launched a similar model
- Access is currently limited to trusted security firms and professionals
- The Trusted Access program is expanding to support thousands of defenders
- Users who pass strict checks can unlock the most advanced features
OpenAI Enters the Digital Defence Race
OpenAI stepping into the cybersecurity space to support ethical hackers in protecting the internet. Just a week earlier, Anthropic released its AI model “Mythos,” which identified thousands of vulnerabilities in software systems. In response, OpenAI developed GPT-5.4-Cyber. This model is trained to detect vulnerable points in the defined code, so they can be patched before malicious actors exploit them to retrieve data or cause systems to fail.
Giving Power to Trusted Experts
OpenAI is also expanding its “Trusted Access for Cyber” program. Launched in February, it now includes thousands of individual experts and hundreds of teams working to secure critical systems, like those used in banking, healthcare, and energy.
To ensure responsible use, access is tiered. Only individuals who pass strict background checks can use the most powerful version. This version has fewer built-in restrictions, allowing deeper analysis of sensitive software to uncover hidden vulnerabilities. This approach helps defenders stay ahead of global cyber threats.
Why It Matters for Australia
For Australia, this development plays an important role in strengthening cybersecurity. As businesses and government systems are frequently attacked, having advanced AI tools to detect vulnerabilities can be valuable. As the major AI companies race to lead in this space, Australian tech teams can use improved tools to protect data and keep important services online.
FAQs
1.What is GPT-5.4-Cyber?
It’s a specialized AI from OpenAI designed to help experts secure systems and websites.
2. Who can use it?
Currently, only vetted security companies, researchers, and trusted teams have access.
3. Is it safe to use?
Yes, access is controlled through a Trusted Access system to ensure only verified users can use its most powerful capabilities.
Follow Inspirepreneur Magazine for daily global business news.