The Digital Age Dilemma
Caught in a whirlwind of technological advancement, we face a critical crossroads: harnessing the immense power of Artificial Intelligence (AI) for good while simultaneously mitigating the potential for catastrophic harm. Artificial Intelligence (AI) has been hailed as the next technological revolution, promising to solve some of humanity’s most pressing problems. However, as AI advances, the risks associated with its misuse grow significantly and could undermine the fabric of society. To ensure that AI benefits humanity, it is essential to prioritise AI ethics and humanity, in its development and deployment.
For thousands of years, humanity has strived to improve quality of life. However, the benefits of technological advancements have often been unevenly distributed, leaving many, especially the world’s poor, marginalized. The digital revolution, if not guided by strong AI ethics and humanity, is likely to exacerbate this inequality.
Data Tsunami: Navigating the Information Age
In this era of rapid technological advancement, data has become the new currency, with 200 zettabytes of data projected to be stored by 2025 (Sausalito, 2024). This tidal wave of information, while fuelling innovation and driving economic growth, also carries significant risks. Cyberattacks are becoming increasingly sophisticated, targeting critical infrastructure and personal data. Data breaches expose sensitive information, leaving individuals and organisations vulnerable to exploitation. The rise of deepfakes, blurs the lines between truth and deception, undermining trust in information and sowing discord in society. As we navigate this data-driven world, it is crucial to develop robust cybersecurity measures, implement strong data privacy regulations, and promote awareness to mitigate the risks and harness the full potential of this digital revolution.
The Human Cost of AI Deception
As AI systems become more sophisticated, there is a growing concern about their potential exploitation for the purposes of manipulating public opinion, disseminating disinformation, and waging cyberwarfare. One alarming example is the use of deepfakes, where AI generates incredibly realistic, yet entirely false, representations of real or imaginary people.
Earlier this year, Arup an engineering company became a victim of this elaborate fraud, to the tune of £20M. The incident involved an employee, who received a video conference invite to discuss urgent, confidential financial transactions. The invitation purported to be from the company’s Chief Financial Officer. Deepfakes have also been used to impersonate politicians, spreading falsehoods among their own supporters. In the US state of New Hampshire, Democrats received calls supposedly from the President Joe Biden, asking them not to vote in the January primary, but to wait until the November general election. The calls were AI-generated, and could have prevented Joe Biden’s name from appearing on the ballot as a Democratic presidential candidate.
AI’s Nightmare Scenario: A Glimpse into the Future
Imagine a world where AI, once a tool for progress, becomes a weapon of control.
A dystopian future is no longer a mere figment of science fiction, and could herald dire consequences:
- Erosion of Individual Liberty: AI-powered surveillance could violate our privacy, suppress dissent, and restrict our freedom of expression.
- Cyber Warfare and Global Disruption: Sophisticated AI-orchestrated cyberattacks could cripple critical infrastructure, destabilise economies, and people facing catastrophic consequences.
- The Rise of Disinformation: Deepfakes inundate the information landscape, undermining trust, tarnishing reputations, inciting anger, shame, loss of funds and in some cases loss of life through bad decision-making.
- Manipulated Minds and Polarized Societies: AI-powered bots, designed to influence and deceive, could control online conversation, amplify extremist views, and fuel social unrest.
A Call to Action: Shaping the Future of AI
To mitigate these risks, it is imperative that policymakers, technologists, and society as a whole work together to develop ethical frameworks and robust regulations for AI. This includes ensuring that AI systems are designed and used in a way that is transparent, accountable, and beneficial to humanity.
While AI offers immense potential, it is essential to approach its development and deployment with caution and foresight. By striking the right balance between innovation, AI ethics and humanity, we can harness the power of AI to create a better future for all.
Given the potential risks and rewards of AI, it is imperative that we take immediate action to shape its future, by applying the following measures:
1. AI Ethics, Humanity, and Governance
(i) Legislation: Laws governing AI applications, especially in critical sectors like healthcare, finance, and law enforcement.
(ii) Ethical: Companies must have ethical AI guidelines, focusing on fairness,transparency, accountability, and prevention of bias. Any breaches should be subject to financial penalties or imprisonment.
(iii) International Collaboration: Global agreements in place that facilitate the tracking and prosecution for harmful use of AI.
2. Ethical and Human-Centered AI Accountability
(i) Explainability: Measures in place that ensure humans understand and can explain the AI system, especially where AI decision-making is involved. For example, the acceptance of candidates for a financial loan can be influenced by their ability to pay back the loan but should not be dependent on personal data, known as “special category data” in General Data Protection Regulation (GDPR).
(ii) Inspection and Control: Regular inspections of AI systems to ensure they are operating within ethical and legal boundaries. Indication of usage and monitoring in company annual reports.
(iii) Liability: Identifying the harmful impact of AI (e.g. human injury or death caused by pacemaker malfunction) and establishing who would be held responsible – the manufacturer, user, or the establishment deploying the AI.
3. Violations of AI Ethics and Human Rights
(i) Deepfakes and Misinformation: Detecting and preventing the misuse of AI, e.g. financial or political gain. AI and human resources in place to identify and thwart images, videos, or news, that are untrue and financial penalties or custodial sentences to be applied to the perpetrators.
(ii) Cybersecurity Threats: Use of non-AI programs along with sophisticated firewalls devised and tested by humans, with additional AI checks are required to ensure systems do not succumb to hacking, phishing, or malware. A crucial factor is the involvement of diverse human resources, not just intellectuals. People from all walks of life think differently, and machines can only learn from the data and algorithms they are provided.
4. Mitigating Bias and Promoting Fairness in AI
(i) Bias: AI systems use algorithms to make predictions or decisions based on patterns identified in training data. These algorithms employ mathematical models to approximate underlying relationships and extrapolate them to new, unseen data. The choice of algorithm can introduce biases, as it may inherently favour certain data patterns or assumptions. Hence, caution must be exercised and bias mitigation actions implemented to ensure fairness.
(ii) Data: Training data may inherently contain biases, for example, 2023 UK government data shows more male fatalities than female fatalities, in the 70+ age group (Reported road casualties great britain, annual report: 2023, 2024). This doesn’t necessarily indicate an expected higher male fatality rate on the road, but could reflect frequency in male and female road usage. Therefore, insurance companies, if using such data, should exercise caution when pricing premiums based on AI modelling.
5. Ethical AI Development
(i) Human-in-the-Loop: In order to prevent harm from autonomous AI systems, we need human beings to check, and if required override automated decision making, to ensure ethical and compassionate outcomes.
(ii) Responsible Innovation: Businesses must encourage AI teams to integrate AI ethics and humanity into their design processes, focusing on transparency, fairness, and accountability.