OpenAI’s Latest Moves to Empower Developers and Tackle Cyber Threats

Artificial Intelligence (AI) has been trending and is poised to bring about substantive changes shortly. Technology holds so much potential to solve most of humanity’s problems. Plans are underway to make AI accessible to users to interact and assist them in solving most social and technical problems.

OpenAI is one of the foremost companies working towards the release of AI algorithms into the
general market. Currently, they have announced some of their AI services that will be of benefit to
developers. To make this quite achievable, they have reduced the prices of some of their APIs. The
GPT-3.5 Turbo is one of the algorithms that has gained a price reduction. The GPT-4 Turbo preview
has been updated and gives more vigorous content moderation capacities. Likewise, the price per
1000 tokens for text-embedding-3-small when compared to text-embedding-ada-002 has dropped
from $0.0001 to 0.00002.

To help developers build safe AI apps, OpenAI has rolled out its advanced content moderation model
yet in text-moderation-007. This will help in identifying potentially harmful texts more precisely than
the previous versions. Developers should also have more control over API keys and visibility into
usage metrics.

UK’s cyber security outfit, National Cyber Security Centre (NCSC) has warned that there will be a
surge of ransomware attacks amid other cyber threats over the next two years. They are concerned
that hackers will use malicious software to encrypt individual files or entire systems and demand
ransom for them. They also worry over the escalation of more phishing attacks which will be used to
trick individuals into giving out sensitive information or clicking on malicious links. These are
problems linked to the advent of generative AI since they can create convincing interactions and generate documents in which phishing red flags will not be uncovered. The center notes that it will be difficult to verify legitimate emails and password reset requests due to generative AI and large language models.

The body advised that developers could improve cyber security through improved attack detection
and system design. They also advised that there should be more focus on research on how
developments in defensive AI solutions can help in reducing cyber security threats.