Qasim Bhatti, CEO of Meta1st examines the emergence of China’s DeepSeek as a viable competitor to ChatGPT and the concerns this raises about security, misinformation and ethical AI deployment
As artificial intelligence (AI) continues its rapid march into the heart of enterprise and consumer technology, emerging new models like China’s DeepSeek are reshaping the competitive AI landscape. While competition drives innovation, it also means we need a deeper scrutiny of the models we choose to trust, integrate and deploy at scale. Not all AI is created equal and the dangers of deploying powerful models without adequate security, oversight or governance have never been more pronounced.
DeepSeek was developed as an open-source alternative to systems like OpenAI’s ChatGPT and it has generated great interest because of its adaptability and flexibility. Yet these very attributes also present the most serious risks to its adoption. As businesses, governments and users across industries increasingly embed AI into decision-making, customer engagement and infrastructure, the importance of evaluating models not just for performance, but for their security posture and ethical footprint, is paramount.
Open AI, open risks
DeepSeek’s development within an open-source framework may on the face of it seem to offer the best of both worlds: technical agility and global collaboration. But the price of that openness is vulnerability. Unlike proprietary models such as ChatGPT, which operate within structured ecosystems governed by ethical AI protocols, DeepSeek lacks centralised safeguards. The potential for misuse is therefore exponentially higher.
One primary risk is adversarial manipulation. Open-source models can be adapted, altered or even poisoned by external actors, either unintentionally or maliciously. In the wrong hands, that flexibility becomes a weapon. Biases can be introduced without scrutiny. Outputs can be subtly warped to fit disinformation narratives. Moreover, because DeepSeek does not enforce a unified compliance framework across implementations, questions arise about its ability to meet regulatory obligations such as GDPR or sector-specific requirements in healthcare, finance or telecommunications.
The absence of built-in moderation adds to the risk. While ChatGPT operates with a defined content policy, fine-tuned moderation layers and enterprise-grade safeguards, DeepSeek relies heavily on users and developers to define the boundaries. This increases the chances of DeepSeek-powered systems being exploited in cyberattacks, leveraged to generate misinformation, or unintentionally violating compliance standards, for example. In a geopolitical
climate where information warfare is a growing concern, the use of unregulated generative AI tools becomes a pressing security issue.
Data protection is another major concern. Open-source models offer fewer guarantees when it comes to safeguarding sensitive inputs. Without the centralised control mechanisms that platforms like ChatGPT use to anonymise or protect user data, DeepSeek leaves the door ajar for breaches, whether by accident or by design.
Governance and ethics must come first
Despite its limitations, DeepSeek represents a significant milestone in global AI development. It shows how fast-moving and ambitious the open-source AI community can be, and it’s a reminder that nations beyond the traditional AI powerhouses of Silicon Valley are building capable alternatives. But the real question isn’t whether DeepSeek can compete on performance – it’s whether it can compete responsibly.
Businesses are understandably drawn to AI tools that offer scalability, multilingual capacity and high availability. But choosing a model is no longer just a technical decision. It is a strategic one, with implications that touch on brand reputation, legal risk, customer trust and long-term sustainability. In this respect, DeepSeek’s lack of corporate oversight and inconsistent governance should be cause for concern.
At Meta1st, we believe AI should only be deployed within frameworks that prioritise ethical integrity, transparency and risk mitigation. That means businesses must ask difficult questions before integrating tools like DeepSeek into their operations. How is bias being detected and mitigated? What protections are in place to prevent adversarial inputs? Who is accountable if the model generates disinformation or discriminatory content? How will regulators interpret your use of a decentralised and potentially non-compliant model?
The responsibility lies not only with developers but with adopters. Choosing an AI model is not dissimilar to selecting a business partner. Trust must be earned, risk must be understood, and safeguards must be provable. As models like DeepSeek are introduced into sensitive sectors, like telecoms, healthcare and national infrastructure, the tolerance for ambiguity diminishes. There is simply too much at stake.
We are entering an era where AI systems will not just assist with workflows, but make decisions that affect lives, liberty and livelihoods. With that evolution comes a burden to ensure these systems are safe, secure and aligned with our values. Proprietary platforms like ChatGPT, while not flawless, have built clear pathways for moderation, auditability and corporate responsibility. DeepSeek, by contrast, remains undefined in its guardrails, and that makes it a risky proposition for businesses without mature internal governance.
This isn’t to suggest that open-source AI has no place in the ecosystem. On the contrary, diverse players must have access to foundational models. But access without accountability creates new vulnerabilities, especially as threat actors, from cyber criminals to state-sponsored
groups, begin to harness AI to scale attacks and influence narratives. When there is no gatekeeper, the system becomes harder to secure.
Ultimately, responsible AI integration requires a posture of vigilance. Organisations must move beyond performance metrics and assess AI through the lens of long-term risk. This means implementing governance frameworks that can detect and respond to anomalies, embedding ethical review processes, and ensuring compliance with evolving regulations. It also means resisting the allure of short-term advantage when it could compromise long-term trust.
DeepSeek’s arrival is a sign of AI’s global maturity. But with greater reach comes greater responsibility. As we evaluate what role it should play in enterprise and public sector environments, we must be prepared to ask not just what it can do – but what it could do in the wrong hands, and whether we are ready for that future.
Inside Telecom provides you with an extensive list of content covering all aspects of the Tech industry. Keep an eye on our Press Releases section to stay informed and updated with our daily articles.